| { | |
| "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": 6.009403705596924, | |
| "learning_rate": 3.125e-07, | |
| "loss": 0.7884, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.019169329073482427, | |
| "grad_norm": 6.14882755279541, | |
| "learning_rate": 6.25e-07, | |
| "loss": 0.8123, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.02875399361022364, | |
| "grad_norm": 6.156098365783691, | |
| "learning_rate": 9.375000000000001e-07, | |
| "loss": 0.8147, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.038338658146964855, | |
| "grad_norm": 5.895145893096924, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.7909, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.04792332268370607, | |
| "grad_norm": 6.2146172523498535, | |
| "learning_rate": 1.5625e-06, | |
| "loss": 0.824, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.05750798722044728, | |
| "grad_norm": 5.762365341186523, | |
| "learning_rate": 1.8750000000000003e-06, | |
| "loss": 0.8087, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0670926517571885, | |
| "grad_norm": 4.7888689041137695, | |
| "learning_rate": 2.1875000000000002e-06, | |
| "loss": 0.774, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.07667731629392971, | |
| "grad_norm": 4.527497291564941, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.737, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.08626198083067092, | |
| "grad_norm": 2.786315679550171, | |
| "learning_rate": 2.8125e-06, | |
| "loss": 0.7153, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.09584664536741214, | |
| "grad_norm": 2.6442039012908936, | |
| "learning_rate": 3.125e-06, | |
| "loss": 0.711, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.10543130990415335, | |
| "grad_norm": 2.196052312850952, | |
| "learning_rate": 3.4375e-06, | |
| "loss": 0.7062, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.11501597444089456, | |
| "grad_norm": 3.406785249710083, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.6862, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.12460063897763578, | |
| "grad_norm": 3.592317581176758, | |
| "learning_rate": 4.0625000000000005e-06, | |
| "loss": 0.6773, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.134185303514377, | |
| "grad_norm": 3.4655966758728027, | |
| "learning_rate": 4.3750000000000005e-06, | |
| "loss": 0.6693, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.14376996805111822, | |
| "grad_norm": 3.5269415378570557, | |
| "learning_rate": 4.6875000000000004e-06, | |
| "loss": 0.664, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.15335463258785942, | |
| "grad_norm": 2.768751621246338, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6127, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.16293929712460065, | |
| "grad_norm": 2.657039165496826, | |
| "learning_rate": 5.3125e-06, | |
| "loss": 0.6004, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.17252396166134185, | |
| "grad_norm": 2.0972790718078613, | |
| "learning_rate": 5.625e-06, | |
| "loss": 0.5971, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.18210862619808307, | |
| "grad_norm": 1.6503219604492188, | |
| "learning_rate": 5.9375e-06, | |
| "loss": 0.5739, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.19169329073482427, | |
| "grad_norm": 1.4270433187484741, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.556, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2012779552715655, | |
| "grad_norm": 1.451823115348816, | |
| "learning_rate": 6.5625e-06, | |
| "loss": 0.5665, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2108626198083067, | |
| "grad_norm": 1.668597936630249, | |
| "learning_rate": 6.875e-06, | |
| "loss": 0.5558, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.22044728434504793, | |
| "grad_norm": 1.7927818298339844, | |
| "learning_rate": 7.1875e-06, | |
| "loss": 0.5801, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.23003194888178913, | |
| "grad_norm": 1.4122098684310913, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.5749, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.23961661341853036, | |
| "grad_norm": 0.9527431130409241, | |
| "learning_rate": 7.8125e-06, | |
| "loss": 0.5295, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.24920127795527156, | |
| "grad_norm": 0.9712406396865845, | |
| "learning_rate": 8.125000000000001e-06, | |
| "loss": 0.5265, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.25878594249201275, | |
| "grad_norm": 1.1668599843978882, | |
| "learning_rate": 8.4375e-06, | |
| "loss": 0.5467, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.268370607028754, | |
| "grad_norm": 0.9096302390098572, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.5272, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2779552715654952, | |
| "grad_norm": 0.8282899856567383, | |
| "learning_rate": 9.0625e-06, | |
| "loss": 0.5265, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.28753993610223644, | |
| "grad_norm": 0.8398028612136841, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 0.5514, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2971246006389776, | |
| "grad_norm": 0.7827602028846741, | |
| "learning_rate": 9.6875e-06, | |
| "loss": 0.4995, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.30670926517571884, | |
| "grad_norm": 0.7462940812110901, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5456, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.31629392971246006, | |
| "grad_norm": 0.7464767694473267, | |
| "learning_rate": 9.999685283773504e-06, | |
| "loss": 0.4866, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3258785942492013, | |
| "grad_norm": 0.7813931703567505, | |
| "learning_rate": 9.998741174712534e-06, | |
| "loss": 0.4905, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.3354632587859425, | |
| "grad_norm": 0.7389088273048401, | |
| "learning_rate": 9.997167791667668e-06, | |
| "loss": 0.5, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3450479233226837, | |
| "grad_norm": 0.6778934597969055, | |
| "learning_rate": 9.994965332706574e-06, | |
| "loss": 0.4928, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3546325878594249, | |
| "grad_norm": 0.6415377855300903, | |
| "learning_rate": 9.992134075089085e-06, | |
| "loss": 0.4866, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.36421725239616615, | |
| "grad_norm": 0.5946885943412781, | |
| "learning_rate": 9.98867437523228e-06, | |
| "loss": 0.494, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.3738019169329074, | |
| "grad_norm": 0.6985592842102051, | |
| "learning_rate": 9.984586668665641e-06, | |
| "loss": 0.4804, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.38338658146964855, | |
| "grad_norm": 0.6060160398483276, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.4642, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3929712460063898, | |
| "grad_norm": 0.5499163866043091, | |
| "learning_rate": 9.974529372743762e-06, | |
| "loss": 0.507, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.402555910543131, | |
| "grad_norm": 0.6382495164871216, | |
| "learning_rate": 9.968561049466214e-06, | |
| "loss": 0.4904, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.41214057507987223, | |
| "grad_norm": 0.6592298746109009, | |
| "learning_rate": 9.961967251474823e-06, | |
| "loss": 0.489, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4217252396166134, | |
| "grad_norm": 0.5930234789848328, | |
| "learning_rate": 9.954748808839675e-06, | |
| "loss": 0.4816, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.43130990415335463, | |
| "grad_norm": 0.5414608716964722, | |
| "learning_rate": 9.946906630265184e-06, | |
| "loss": 0.4842, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.44089456869009586, | |
| "grad_norm": 0.5305256247520447, | |
| "learning_rate": 9.938441702975689e-06, | |
| "loss": 0.4786, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4504792332268371, | |
| "grad_norm": 0.8502335548400879, | |
| "learning_rate": 9.92935509259118e-06, | |
| "loss": 0.4924, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.46006389776357826, | |
| "grad_norm": 0.6727577447891235, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.4859, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4696485623003195, | |
| "grad_norm": 0.6252870559692383, | |
| "learning_rate": 9.909321476180594e-06, | |
| "loss": 0.4837, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.4792332268370607, | |
| "grad_norm": 0.6533059477806091, | |
| "learning_rate": 9.898376992116179e-06, | |
| "loss": 0.4695, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.48881789137380194, | |
| "grad_norm": 0.7547851204872131, | |
| "learning_rate": 9.886815868562596e-06, | |
| "loss": 0.4633, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4984025559105431, | |
| "grad_norm": 0.6159355640411377, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.4716, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5079872204472844, | |
| "grad_norm": 0.5551995038986206, | |
| "learning_rate": 9.861849601988384e-06, | |
| "loss": 0.4647, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5175718849840255, | |
| "grad_norm": 0.6664530038833618, | |
| "learning_rate": 9.848447601883436e-06, | |
| "loss": 0.4676, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5271565495207667, | |
| "grad_norm": 0.6922640800476074, | |
| "learning_rate": 9.834435247725032e-06, | |
| "loss": 0.4656, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.536741214057508, | |
| "grad_norm": 0.5392067432403564, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.4611, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5463258785942492, | |
| "grad_norm": 0.5382002592086792, | |
| "learning_rate": 9.804586609725499e-06, | |
| "loss": 0.4414, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5559105431309904, | |
| "grad_norm": 0.6740583777427673, | |
| "learning_rate": 9.788754083424654e-06, | |
| "loss": 0.465, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5654952076677316, | |
| "grad_norm": 0.5912691950798035, | |
| "learning_rate": 9.772318717677905e-06, | |
| "loss": 0.4549, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.5750798722044729, | |
| "grad_norm": 0.517918586730957, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.481, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5846645367412141, | |
| "grad_norm": 0.6886275410652161, | |
| "learning_rate": 9.737647819437645e-06, | |
| "loss": 0.4552, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5942492012779552, | |
| "grad_norm": 0.5352553129196167, | |
| "learning_rate": 9.719416651541839e-06, | |
| "loss": 0.4422, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6038338658146964, | |
| "grad_norm": 0.5464288592338562, | |
| "learning_rate": 9.700591372846096e-06, | |
| "loss": 0.482, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6134185303514377, | |
| "grad_norm": 0.5277808904647827, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.4552, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.6230031948881789, | |
| "grad_norm": 0.6267250180244446, | |
| "learning_rate": 9.661168036940071e-06, | |
| "loss": 0.459, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6325878594249201, | |
| "grad_norm": 0.486532598733902, | |
| "learning_rate": 9.640574942595195e-06, | |
| "loss": 0.4472, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6421725239616614, | |
| "grad_norm": 0.6028465032577515, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.4269, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6517571884984026, | |
| "grad_norm": 0.5778775811195374, | |
| "learning_rate": 9.597638862757255e-06, | |
| "loss": 0.4835, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6613418530351438, | |
| "grad_norm": 0.5969807505607605, | |
| "learning_rate": 9.5753012823366e-06, | |
| "loss": 0.4626, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.670926517571885, | |
| "grad_norm": 0.5960472226142883, | |
| "learning_rate": 9.552387733294081e-06, | |
| "loss": 0.4466, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6805111821086262, | |
| "grad_norm": 0.5721907615661621, | |
| "learning_rate": 9.528901100135971e-06, | |
| "loss": 0.4532, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6900958466453674, | |
| "grad_norm": 0.5902190804481506, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.4516, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.6996805111821086, | |
| "grad_norm": 0.61021488904953, | |
| "learning_rate": 9.480220479843627e-06, | |
| "loss": 0.4353, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7092651757188498, | |
| "grad_norm": 0.5561513304710388, | |
| "learning_rate": 9.45503262094184e-06, | |
| "loss": 0.4384, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.7188498402555911, | |
| "grad_norm": 0.610275387763977, | |
| "learning_rate": 9.4292839336179e-06, | |
| "loss": 0.4252, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7284345047923323, | |
| "grad_norm": 0.5793384909629822, | |
| "learning_rate": 9.40297765928369e-06, | |
| "loss": 0.4489, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.7380191693290735, | |
| "grad_norm": 0.5056276321411133, | |
| "learning_rate": 9.376117109543769e-06, | |
| "loss": 0.4526, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.7476038338658147, | |
| "grad_norm": 0.4884008467197418, | |
| "learning_rate": 9.348705665778479e-06, | |
| "loss": 0.4515, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7571884984025559, | |
| "grad_norm": 0.5363313555717468, | |
| "learning_rate": 9.320746778718274e-06, | |
| "loss": 0.4443, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.7667731629392971, | |
| "grad_norm": 0.5665227174758911, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.4235, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7763578274760383, | |
| "grad_norm": 0.5609092712402344, | |
| "learning_rate": 9.263200821770462e-06, | |
| "loss": 0.4651, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7859424920127795, | |
| "grad_norm": 0.5751113891601562, | |
| "learning_rate": 9.233620996141421e-06, | |
| "loss": 0.453, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7955271565495208, | |
| "grad_norm": 0.6094540953636169, | |
| "learning_rate": 9.203508214822652e-06, | |
| "loss": 0.4462, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.805111821086262, | |
| "grad_norm": 0.5433350205421448, | |
| "learning_rate": 9.172866268606514e-06, | |
| "loss": 0.4407, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.8146964856230032, | |
| "grad_norm": 0.6528311371803284, | |
| "learning_rate": 9.141699014900084e-06, | |
| "loss": 0.4385, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8242811501597445, | |
| "grad_norm": 0.6348373293876648, | |
| "learning_rate": 9.110010377239552e-06, | |
| "loss": 0.4671, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.8338658146964856, | |
| "grad_norm": 0.6578677892684937, | |
| "learning_rate": 9.077804344796302e-06, | |
| "loss": 0.44, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.8434504792332268, | |
| "grad_norm": 0.6142277717590332, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.4682, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.853035143769968, | |
| "grad_norm": 0.5283975601196289, | |
| "learning_rate": 9.011856377401891e-06, | |
| "loss": 0.4392, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.8626198083067093, | |
| "grad_norm": 0.7270615100860596, | |
| "learning_rate": 8.978122744408905e-06, | |
| "loss": 0.4346, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8722044728434505, | |
| "grad_norm": 0.6391173005104065, | |
| "learning_rate": 8.943888319504456e-06, | |
| "loss": 0.4386, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8817891373801917, | |
| "grad_norm": 0.5456526279449463, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.4393, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8913738019169329, | |
| "grad_norm": 0.677370011806488, | |
| "learning_rate": 8.873934395068006e-06, | |
| "loss": 0.4614, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.9009584664536742, | |
| "grad_norm": 0.5470128059387207, | |
| "learning_rate": 8.838223701790057e-06, | |
| "loss": 0.4399, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.9105431309904153, | |
| "grad_norm": 0.5786166191101074, | |
| "learning_rate": 8.802029828000157e-06, | |
| "loss": 0.4462, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9201277955271565, | |
| "grad_norm": 0.5874597430229187, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.4644, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.9297124600638977, | |
| "grad_norm": 0.5426689386367798, | |
| "learning_rate": 8.728210824415829e-06, | |
| "loss": 0.4381, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.939297124600639, | |
| "grad_norm": 0.5549809336662292, | |
| "learning_rate": 8.690594987436705e-06, | |
| "loss": 0.4283, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.9488817891373802, | |
| "grad_norm": 0.6467133164405823, | |
| "learning_rate": 8.652514554406388e-06, | |
| "loss": 0.4572, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.9584664536741214, | |
| "grad_norm": 0.5148814916610718, | |
| "learning_rate": 8.613974319136959e-06, | |
| "loss": 0.4381, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9680511182108626, | |
| "grad_norm": 0.4909837245941162, | |
| "learning_rate": 8.574979133323378e-06, | |
| "loss": 0.4543, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9776357827476039, | |
| "grad_norm": 0.6221508383750916, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.4601, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.987220447284345, | |
| "grad_norm": 0.615908145904541, | |
| "learning_rate": 8.495643602586287e-06, | |
| "loss": 0.4398, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9968051118210862, | |
| "grad_norm": 0.6509765386581421, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.4556, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.0063897763578276, | |
| "grad_norm": 0.5689537525177002, | |
| "learning_rate": 8.414547910024035e-06, | |
| "loss": 0.3969, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.0159744408945688, | |
| "grad_norm": 0.6029090881347656, | |
| "learning_rate": 8.373352729660373e-06, | |
| "loss": 0.4141, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.0255591054313098, | |
| "grad_norm": 0.5475075840950012, | |
| "learning_rate": 8.331732889760021e-06, | |
| "loss": 0.4064, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.035143769968051, | |
| "grad_norm": 0.5231770277023315, | |
| "learning_rate": 8.289693629698564e-06, | |
| "loss": 0.4148, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.0447284345047922, | |
| "grad_norm": 0.5678158402442932, | |
| "learning_rate": 8.247240241650918e-06, | |
| "loss": 0.3855, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.0543130990415335, | |
| "grad_norm": 0.5159103870391846, | |
| "learning_rate": 8.204378069925121e-06, | |
| "loss": 0.404, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0638977635782747, | |
| "grad_norm": 0.5033451914787292, | |
| "learning_rate": 8.16111251028955e-06, | |
| "loss": 0.3802, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.073482428115016, | |
| "grad_norm": 0.593724250793457, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.4093, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.0830670926517572, | |
| "grad_norm": 0.5523594617843628, | |
| "learning_rate": 8.073393063582386e-06, | |
| "loss": 0.4117, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0926517571884984, | |
| "grad_norm": 0.5447916388511658, | |
| "learning_rate": 8.0289502192041e-06, | |
| "loss": 0.4056, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.1022364217252396, | |
| "grad_norm": 0.5257798433303833, | |
| "learning_rate": 7.984126070912519e-06, | |
| "loss": 0.4015, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.1118210862619808, | |
| "grad_norm": 0.579169511795044, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.373, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.121405750798722, | |
| "grad_norm": 0.5695212483406067, | |
| "learning_rate": 7.89335648089903e-06, | |
| "loss": 0.379, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.1309904153354633, | |
| "grad_norm": 0.5016732215881348, | |
| "learning_rate": 7.84742246584226e-06, | |
| "loss": 0.411, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.1405750798722045, | |
| "grad_norm": 0.5948535799980164, | |
| "learning_rate": 7.801129998764014e-06, | |
| "loss": 0.3903, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.1501597444089458, | |
| "grad_norm": 0.5826632380485535, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.377, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.159744408945687, | |
| "grad_norm": 0.5407915711402893, | |
| "learning_rate": 7.70749306331863e-06, | |
| "loss": 0.3975, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.1693290734824282, | |
| "grad_norm": 0.5077962279319763, | |
| "learning_rate": 7.660160382576683e-06, | |
| "loss": 0.4125, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.1789137380191694, | |
| "grad_norm": 0.5673754811286926, | |
| "learning_rate": 7.612492823579744e-06, | |
| "loss": 0.3853, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.1884984025559104, | |
| "grad_norm": 0.5954229831695557, | |
| "learning_rate": 7.564496387029532e-06, | |
| "loss": 0.4016, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.1980830670926517, | |
| "grad_norm": 0.5240495800971985, | |
| "learning_rate": 7.516177115029002e-06, | |
| "loss": 0.4077, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.207667731629393, | |
| "grad_norm": 0.5433353781700134, | |
| "learning_rate": 7.467541090321735e-06, | |
| "loss": 0.3842, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.2172523961661341, | |
| "grad_norm": 0.5105989575386047, | |
| "learning_rate": 7.4185944355261996e-06, | |
| "loss": 0.3848, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.2268370607028753, | |
| "grad_norm": 0.5643378496170044, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.4038, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.2364217252396166, | |
| "grad_norm": 0.5455153584480286, | |
| "learning_rate": 7.319793920889171e-06, | |
| "loss": 0.396, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.2460063897763578, | |
| "grad_norm": 0.5321876406669617, | |
| "learning_rate": 7.269952498697734e-06, | |
| "loss": 0.389, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.255591054313099, | |
| "grad_norm": 0.58072829246521, | |
| "learning_rate": 7.219825320152411e-06, | |
| "loss": 0.3873, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.2651757188498403, | |
| "grad_norm": 0.5760256052017212, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.3905, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.2747603833865815, | |
| "grad_norm": 0.6082053184509277, | |
| "learning_rate": 7.118738970516944e-06, | |
| "loss": 0.3793, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.2843450479233227, | |
| "grad_norm": 0.5610427856445312, | |
| "learning_rate": 7.067792524832604e-06, | |
| "loss": 0.3942, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.293929712460064, | |
| "grad_norm": 0.5390857458114624, | |
| "learning_rate": 7.016585772004026e-06, | |
| "loss": 0.3816, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.3035143769968052, | |
| "grad_norm": 0.6185169816017151, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.3908, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.3130990415335464, | |
| "grad_norm": 0.4800776541233063, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.4012, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.3226837060702876, | |
| "grad_norm": 0.4752432703971863, | |
| "learning_rate": 6.8614682920097265e-06, | |
| "loss": 0.3831, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.3322683706070286, | |
| "grad_norm": 0.4728682041168213, | |
| "learning_rate": 6.809285088483361e-06, | |
| "loss": 0.3819, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.34185303514377, | |
| "grad_norm": 0.5230915546417236, | |
| "learning_rate": 6.7568741204067145e-06, | |
| "loss": 0.4063, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.351437699680511, | |
| "grad_norm": 0.4786807894706726, | |
| "learning_rate": 6.704241985612625e-06, | |
| "loss": 0.3885, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.3610223642172525, | |
| "grad_norm": 0.5374184250831604, | |
| "learning_rate": 6.651395309775837e-06, | |
| "loss": 0.3562, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.3706070287539935, | |
| "grad_norm": 0.4862798750400543, | |
| "learning_rate": 6.598340745578908e-06, | |
| "loss": 0.3751, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.3801916932907348, | |
| "grad_norm": 0.49519556760787964, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.4145, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.389776357827476, | |
| "grad_norm": 0.5167025923728943, | |
| "learning_rate": 6.491634692845781e-06, | |
| "loss": 0.3769, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.3993610223642172, | |
| "grad_norm": 0.4625193774700165, | |
| "learning_rate": 6.437996637160086e-06, | |
| "loss": 0.365, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.4089456869009584, | |
| "grad_norm": 0.4104255735874176, | |
| "learning_rate": 6.384177557124247e-06, | |
| "loss": 0.3834, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.4185303514376997, | |
| "grad_norm": 0.42858555912971497, | |
| "learning_rate": 6.330184227833376e-06, | |
| "loss": 0.3932, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.428115015974441, | |
| "grad_norm": 0.47697922587394714, | |
| "learning_rate": 6.276023446318214e-06, | |
| "loss": 0.4063, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.4376996805111821, | |
| "grad_norm": 0.43023285269737244, | |
| "learning_rate": 6.2217020306894705e-06, | |
| "loss": 0.3907, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4472843450479234, | |
| "grad_norm": 0.4481910169124603, | |
| "learning_rate": 6.1672268192795285e-06, | |
| "loss": 0.3953, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.4568690095846646, | |
| "grad_norm": 0.4804971218109131, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.4112, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.4664536741214058, | |
| "grad_norm": 0.4408077001571655, | |
| "learning_rate": 6.057842458386315e-06, | |
| "loss": 0.4019, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.476038338658147, | |
| "grad_norm": 0.4253651201725006, | |
| "learning_rate": 6.002947078916365e-06, | |
| "loss": 0.3874, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.4856230031948883, | |
| "grad_norm": 0.47946813702583313, | |
| "learning_rate": 5.947925441958393e-06, | |
| "loss": 0.3891, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.4952076677316293, | |
| "grad_norm": 0.45089444518089294, | |
| "learning_rate": 5.892784473993184e-06, | |
| "loss": 0.3914, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.5047923322683707, | |
| "grad_norm": 0.4402899444103241, | |
| "learning_rate": 5.837531116523683e-06, | |
| "loss": 0.403, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.5143769968051117, | |
| "grad_norm": 0.4637235999107361, | |
| "learning_rate": 5.782172325201155e-06, | |
| "loss": 0.3895, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.5239616613418532, | |
| "grad_norm": 0.4778022766113281, | |
| "learning_rate": 5.726715068949564e-06, | |
| "loss": 0.3818, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.5335463258785942, | |
| "grad_norm": 0.44513994455337524, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.3911, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5431309904153354, | |
| "grad_norm": 0.4892440736293793, | |
| "learning_rate": 5.615533098453215e-06, | |
| "loss": 0.3924, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.5527156549520766, | |
| "grad_norm": 0.4401436448097229, | |
| "learning_rate": 5.559822380516539e-06, | |
| "loss": 0.3833, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.5623003194888179, | |
| "grad_norm": 0.43897783756256104, | |
| "learning_rate": 5.504041188505022e-06, | |
| "loss": 0.3953, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.571884984025559, | |
| "grad_norm": 0.44918692111968994, | |
| "learning_rate": 5.448196544517168e-06, | |
| "loss": 0.3963, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.5814696485623003, | |
| "grad_norm": 0.5550315380096436, | |
| "learning_rate": 5.392295478639226e-06, | |
| "loss": 0.3909, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.5910543130990416, | |
| "grad_norm": 0.5739699006080627, | |
| "learning_rate": 5.336345028060199e-06, | |
| "loss": 0.4128, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.6006389776357828, | |
| "grad_norm": 0.47378113865852356, | |
| "learning_rate": 5.2803522361859596e-06, | |
| "loss": 0.3823, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.610223642172524, | |
| "grad_norm": 0.4881516098976135, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.402, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.619808306709265, | |
| "grad_norm": 0.5072901248931885, | |
| "learning_rate": 5.168267827938971e-06, | |
| "loss": 0.3923, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.6293929712460065, | |
| "grad_norm": 0.5656172037124634, | |
| "learning_rate": 5.112190321479026e-06, | |
| "loss": 0.396, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6389776357827475, | |
| "grad_norm": 0.437597393989563, | |
| "learning_rate": 5.05609869177323e-06, | |
| "loss": 0.3849, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.648562300319489, | |
| "grad_norm": 0.47302526235580444, | |
| "learning_rate": 5e-06, | |
| "loss": 0.3811, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.65814696485623, | |
| "grad_norm": 0.5107726454734802, | |
| "learning_rate": 4.943901308226771e-06, | |
| "loss": 0.3683, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.6677316293929714, | |
| "grad_norm": 0.46889111399650574, | |
| "learning_rate": 4.887809678520976e-06, | |
| "loss": 0.381, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.6773162939297124, | |
| "grad_norm": 0.5070697665214539, | |
| "learning_rate": 4.831732172061032e-06, | |
| "loss": 0.3791, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.6869009584664538, | |
| "grad_norm": 0.4947984218597412, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.3578, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.6964856230031948, | |
| "grad_norm": 0.45956239104270935, | |
| "learning_rate": 4.719647763814041e-06, | |
| "loss": 0.3853, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.706070287539936, | |
| "grad_norm": 0.4792458415031433, | |
| "learning_rate": 4.663654971939802e-06, | |
| "loss": 0.3978, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.7156549520766773, | |
| "grad_norm": 0.45664891600608826, | |
| "learning_rate": 4.6077045213607765e-06, | |
| "loss": 0.3769, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.7252396166134185, | |
| "grad_norm": 0.4743683934211731, | |
| "learning_rate": 4.551803455482833e-06, | |
| "loss": 0.395, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.7348242811501597, | |
| "grad_norm": 0.4895039200782776, | |
| "learning_rate": 4.4959588114949785e-06, | |
| "loss": 0.3829, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.744408945686901, | |
| "grad_norm": 0.4933681786060333, | |
| "learning_rate": 4.4401776194834615e-06, | |
| "loss": 0.3861, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.7539936102236422, | |
| "grad_norm": 0.4807256758213043, | |
| "learning_rate": 4.384466901546786e-06, | |
| "loss": 0.3909, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.7635782747603834, | |
| "grad_norm": 0.4260290563106537, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.3955, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.7731629392971247, | |
| "grad_norm": 0.4630730152130127, | |
| "learning_rate": 4.273284931050438e-06, | |
| "loss": 0.3991, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.7827476038338657, | |
| "grad_norm": 0.506591796875, | |
| "learning_rate": 4.217827674798845e-06, | |
| "loss": 0.3942, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.792332268370607, | |
| "grad_norm": 0.4698037803173065, | |
| "learning_rate": 4.162468883476319e-06, | |
| "loss": 0.3814, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.8019169329073481, | |
| "grad_norm": 0.44776004552841187, | |
| "learning_rate": 4.107215526006818e-06, | |
| "loss": 0.425, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.8115015974440896, | |
| "grad_norm": 0.4561665952205658, | |
| "learning_rate": 4.052074558041608e-06, | |
| "loss": 0.3965, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.8210862619808306, | |
| "grad_norm": 0.4227313697338104, | |
| "learning_rate": 3.997052921083637e-06, | |
| "loss": 0.3812, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.830670926517572, | |
| "grad_norm": 0.4525080621242523, | |
| "learning_rate": 3.9421575416136866e-06, | |
| "loss": 0.3904, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.840255591054313, | |
| "grad_norm": 0.44339868426322937, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.411, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.8498402555910545, | |
| "grad_norm": 0.41708871722221375, | |
| "learning_rate": 3.832773180720475e-06, | |
| "loss": 0.3628, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.8594249201277955, | |
| "grad_norm": 0.4562470018863678, | |
| "learning_rate": 3.778297969310529e-06, | |
| "loss": 0.4054, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.8690095846645367, | |
| "grad_norm": 0.4357210099697113, | |
| "learning_rate": 3.723976553681787e-06, | |
| "loss": 0.3843, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.878594249201278, | |
| "grad_norm": 0.4327453076839447, | |
| "learning_rate": 3.669815772166625e-06, | |
| "loss": 0.4009, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.8881789137380192, | |
| "grad_norm": 0.48004162311553955, | |
| "learning_rate": 3.6158224428757538e-06, | |
| "loss": 0.406, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.8977635782747604, | |
| "grad_norm": 0.43024709820747375, | |
| "learning_rate": 3.562003362839914e-06, | |
| "loss": 0.3711, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.9073482428115016, | |
| "grad_norm": 0.42549964785575867, | |
| "learning_rate": 3.50836530715422e-06, | |
| "loss": 0.4026, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.9169329073482428, | |
| "grad_norm": 0.4745008051395416, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.4173, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.926517571884984, | |
| "grad_norm": 0.4466969668865204, | |
| "learning_rate": 3.4016592544210937e-06, | |
| "loss": 0.3825, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.9361022364217253, | |
| "grad_norm": 0.4486326575279236, | |
| "learning_rate": 3.3486046902241663e-06, | |
| "loss": 0.4089, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.9456869009584663, | |
| "grad_norm": 0.39550328254699707, | |
| "learning_rate": 3.295758014387375e-06, | |
| "loss": 0.385, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.9552715654952078, | |
| "grad_norm": 0.4342930018901825, | |
| "learning_rate": 3.2431258795932863e-06, | |
| "loss": 0.3903, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.9648562300319488, | |
| "grad_norm": 0.4231680631637573, | |
| "learning_rate": 3.1907149115166403e-06, | |
| "loss": 0.3809, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.9744408945686902, | |
| "grad_norm": 0.4089261293411255, | |
| "learning_rate": 3.1385317079902743e-06, | |
| "loss": 0.4005, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.9840255591054312, | |
| "grad_norm": 0.4242880940437317, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.3893, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.9936102236421727, | |
| "grad_norm": 0.44006791710853577, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.3569, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.0031948881789137, | |
| "grad_norm": 0.4005739986896515, | |
| "learning_rate": 2.9834142279959754e-06, | |
| "loss": 0.3688, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.012779552715655, | |
| "grad_norm": 0.46779128909111023, | |
| "learning_rate": 2.932207475167398e-06, | |
| "loss": 0.3468, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.022364217252396, | |
| "grad_norm": 0.5104973912239075, | |
| "learning_rate": 2.8812610294830568e-06, | |
| "loss": 0.3356, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.0319488817891376, | |
| "grad_norm": 0.41405051946640015, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.3633, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.0415335463258786, | |
| "grad_norm": 0.42602765560150146, | |
| "learning_rate": 2.7801746798475905e-06, | |
| "loss": 0.3453, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.0511182108626196, | |
| "grad_norm": 0.3947254419326782, | |
| "learning_rate": 2.7300475013022666e-06, | |
| "loss": 0.3232, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.060702875399361, | |
| "grad_norm": 0.43924054503440857, | |
| "learning_rate": 2.6802060791108304e-06, | |
| "loss": 0.3385, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.070287539936102, | |
| "grad_norm": 0.46914729475975037, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.3629, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.0798722044728435, | |
| "grad_norm": 0.4117669463157654, | |
| "learning_rate": 2.5814055644738013e-06, | |
| "loss": 0.3418, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.0894568690095845, | |
| "grad_norm": 0.41317105293273926, | |
| "learning_rate": 2.532458909678266e-06, | |
| "loss": 0.3421, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.099041533546326, | |
| "grad_norm": 0.4393919110298157, | |
| "learning_rate": 2.483822884971e-06, | |
| "loss": 0.3529, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.108626198083067, | |
| "grad_norm": 0.46012818813323975, | |
| "learning_rate": 2.43550361297047e-06, | |
| "loss": 0.3292, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.1182108626198084, | |
| "grad_norm": 0.4242909252643585, | |
| "learning_rate": 2.387507176420256e-06, | |
| "loss": 0.3327, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.1277955271565494, | |
| "grad_norm": 0.41850873827934265, | |
| "learning_rate": 2.339839617423318e-06, | |
| "loss": 0.3478, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.137380191693291, | |
| "grad_norm": 0.450730562210083, | |
| "learning_rate": 2.2925069366813718e-06, | |
| "loss": 0.359, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.146964856230032, | |
| "grad_norm": 0.4045020639896393, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.3193, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.1565495207667733, | |
| "grad_norm": 0.4146140515804291, | |
| "learning_rate": 2.1988700012359865e-06, | |
| "loss": 0.3534, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.1661341853035143, | |
| "grad_norm": 0.4432181119918823, | |
| "learning_rate": 2.1525775341577404e-06, | |
| "loss": 0.3595, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.1757188498402558, | |
| "grad_norm": 0.4246091842651367, | |
| "learning_rate": 2.1066435191009717e-06, | |
| "loss": 0.349, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.1853035143769968, | |
| "grad_norm": 0.40804770588874817, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.3481, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.194888178913738, | |
| "grad_norm": 0.4196283221244812, | |
| "learning_rate": 2.0158739290874822e-06, | |
| "loss": 0.3539, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.2044728434504792, | |
| "grad_norm": 0.38730764389038086, | |
| "learning_rate": 1.971049780795901e-06, | |
| "loss": 0.3637, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.2140575079872207, | |
| "grad_norm": 0.39496931433677673, | |
| "learning_rate": 1.9266069364176144e-06, | |
| "loss": 0.3391, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.2236421725239617, | |
| "grad_norm": 0.4084094166755676, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.3245, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.2332268370607027, | |
| "grad_norm": 0.39960479736328125, | |
| "learning_rate": 1.838887489710452e-06, | |
| "loss": 0.3564, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.242811501597444, | |
| "grad_norm": 0.3978304862976074, | |
| "learning_rate": 1.7956219300748796e-06, | |
| "loss": 0.3457, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.252396166134185, | |
| "grad_norm": 0.39459142088890076, | |
| "learning_rate": 1.7527597583490825e-06, | |
| "loss": 0.3503, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.2619808306709266, | |
| "grad_norm": 0.41241538524627686, | |
| "learning_rate": 1.7103063703014372e-06, | |
| "loss": 0.3496, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.2715654952076676, | |
| "grad_norm": 0.40372639894485474, | |
| "learning_rate": 1.6682671102399806e-06, | |
| "loss": 0.3379, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.281150159744409, | |
| "grad_norm": 0.371053546667099, | |
| "learning_rate": 1.6266472703396286e-06, | |
| "loss": 0.3419, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.29073482428115, | |
| "grad_norm": 0.42974787950515747, | |
| "learning_rate": 1.5854520899759656e-06, | |
| "loss": 0.3622, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.3003194888178915, | |
| "grad_norm": 0.3963543474674225, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.3741, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.3099041533546325, | |
| "grad_norm": 0.37718451023101807, | |
| "learning_rate": 1.5043563974137132e-06, | |
| "loss": 0.337, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.319488817891374, | |
| "grad_norm": 0.3763139843940735, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.3303, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.329073482428115, | |
| "grad_norm": 0.3658771216869354, | |
| "learning_rate": 1.4250208666766235e-06, | |
| "loss": 0.3278, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.3386581469648564, | |
| "grad_norm": 0.38942959904670715, | |
| "learning_rate": 1.3860256808630429e-06, | |
| "loss": 0.3611, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.3482428115015974, | |
| "grad_norm": 0.3814983069896698, | |
| "learning_rate": 1.3474854455936126e-06, | |
| "loss": 0.3266, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.357827476038339, | |
| "grad_norm": 0.41327863931655884, | |
| "learning_rate": 1.3094050125632973e-06, | |
| "loss": 0.34, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.36741214057508, | |
| "grad_norm": 0.41362810134887695, | |
| "learning_rate": 1.2717891755841722e-06, | |
| "loss": 0.3297, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.376996805111821, | |
| "grad_norm": 0.41169583797454834, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.3354, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.3865814696485623, | |
| "grad_norm": 0.37766772508621216, | |
| "learning_rate": 1.1979701719998454e-06, | |
| "loss": 0.3657, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.3961661341853033, | |
| "grad_norm": 0.38726282119750977, | |
| "learning_rate": 1.1617762982099446e-06, | |
| "loss": 0.35, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.405750798722045, | |
| "grad_norm": 0.3911564350128174, | |
| "learning_rate": 1.1260656049319957e-06, | |
| "loss": 0.3545, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.415335463258786, | |
| "grad_norm": 0.413602739572525, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.3842, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.4249201277955272, | |
| "grad_norm": 0.3559524416923523, | |
| "learning_rate": 1.0561116804955451e-06, | |
| "loss": 0.345, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.4345047923322682, | |
| "grad_norm": 0.36401858925819397, | |
| "learning_rate": 1.0218772555910955e-06, | |
| "loss": 0.3386, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.4440894568690097, | |
| "grad_norm": 0.38348931074142456, | |
| "learning_rate": 9.881436225981107e-07, | |
| "loss": 0.3344, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.4536741214057507, | |
| "grad_norm": 0.3621443510055542, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.3199, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.463258785942492, | |
| "grad_norm": 0.38067299127578735, | |
| "learning_rate": 9.221956552036992e-07, | |
| "loss": 0.355, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.472843450479233, | |
| "grad_norm": 0.3758901059627533, | |
| "learning_rate": 8.899896227604509e-07, | |
| "loss": 0.3365, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.4824281150159746, | |
| "grad_norm": 0.36369767785072327, | |
| "learning_rate": 8.58300985099918e-07, | |
| "loss": 0.3234, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.4920127795527156, | |
| "grad_norm": 0.40129390358924866, | |
| "learning_rate": 8.271337313934869e-07, | |
| "loss": 0.3562, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.501597444089457, | |
| "grad_norm": 0.37340959906578064, | |
| "learning_rate": 7.964917851773496e-07, | |
| "loss": 0.3304, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.511182108626198, | |
| "grad_norm": 0.35955309867858887, | |
| "learning_rate": 7.663790038585794e-07, | |
| "loss": 0.3373, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.520766773162939, | |
| "grad_norm": 0.3691595792770386, | |
| "learning_rate": 7.367991782295392e-07, | |
| "loss": 0.3463, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.5303514376996805, | |
| "grad_norm": 0.35693010687828064, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.3569, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.539936102236422, | |
| "grad_norm": 0.33215582370758057, | |
| "learning_rate": 6.792532212817271e-07, | |
| "loss": 0.3686, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.549520766773163, | |
| "grad_norm": 0.3821455240249634, | |
| "learning_rate": 6.512943342215234e-07, | |
| "loss": 0.3324, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.559105431309904, | |
| "grad_norm": 0.37147057056427, | |
| "learning_rate": 6.238828904562316e-07, | |
| "loss": 0.3527, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.5686900958466454, | |
| "grad_norm": 0.35479873418807983, | |
| "learning_rate": 5.9702234071631e-07, | |
| "loss": 0.3328, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.5782747603833864, | |
| "grad_norm": 0.3780212998390198, | |
| "learning_rate": 5.707160663821009e-07, | |
| "loss": 0.3535, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.587859424920128, | |
| "grad_norm": 0.3386017978191376, | |
| "learning_rate": 5.449673790581611e-07, | |
| "loss": 0.3701, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.597444089456869, | |
| "grad_norm": 0.3443048596382141, | |
| "learning_rate": 5.197795201563744e-07, | |
| "loss": 0.3678, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.6070287539936103, | |
| "grad_norm": 0.3585163950920105, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.3279, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.6166134185303513, | |
| "grad_norm": 0.37369778752326965, | |
| "learning_rate": 4.710988998640298e-07, | |
| "loss": 0.3415, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.626198083067093, | |
| "grad_norm": 0.35122641921043396, | |
| "learning_rate": 4.4761226670592074e-07, | |
| "loss": 0.3187, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.635782747603834, | |
| "grad_norm": 0.37191906571388245, | |
| "learning_rate": 4.2469871766340096e-07, | |
| "loss": 0.3601, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.6453674121405752, | |
| "grad_norm": 0.3587002754211426, | |
| "learning_rate": 4.0236113724274716e-07, | |
| "loss": 0.343, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.6549520766773163, | |
| "grad_norm": 0.36429116129875183, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.321, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.6645367412140573, | |
| "grad_norm": 0.37538281083106995, | |
| "learning_rate": 3.5942505740480583e-07, | |
| "loss": 0.3472, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.6741214057507987, | |
| "grad_norm": 0.3428320288658142, | |
| "learning_rate": 3.3883196305992906e-07, | |
| "loss": 0.3541, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.68370607028754, | |
| "grad_norm": 0.38002461194992065, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.343, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.693290734824281, | |
| "grad_norm": 0.3926028609275818, | |
| "learning_rate": 2.9940862715390483e-07, | |
| "loss": 0.3266, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.702875399361022, | |
| "grad_norm": 0.3759942054748535, | |
| "learning_rate": 2.8058334845816214e-07, | |
| "loss": 0.3611, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.7124600638977636, | |
| "grad_norm": 0.3495001494884491, | |
| "learning_rate": 2.6235218056235633e-07, | |
| "loss": 0.3415, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.722044728434505, | |
| "grad_norm": 0.3587774932384491, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 0.3673, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.731629392971246, | |
| "grad_norm": 0.3531428873538971, | |
| "learning_rate": 2.276812823220964e-07, | |
| "loss": 0.3675, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.741214057507987, | |
| "grad_norm": 0.3556656241416931, | |
| "learning_rate": 2.1124591657534776e-07, | |
| "loss": 0.3439, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.7507987220447285, | |
| "grad_norm": 0.3901713192462921, | |
| "learning_rate": 1.9541339027450256e-07, | |
| "loss": 0.3376, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.7603833865814695, | |
| "grad_norm": 0.3528308868408203, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.3697, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.769968051118211, | |
| "grad_norm": 0.35200899839401245, | |
| "learning_rate": 1.6556475227496816e-07, | |
| "loss": 0.3419, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.779552715654952, | |
| "grad_norm": 0.3826752305030823, | |
| "learning_rate": 1.5155239811656562e-07, | |
| "loss": 0.3325, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.7891373801916934, | |
| "grad_norm": 0.3455723524093628, | |
| "learning_rate": 1.3815039801161723e-07, | |
| "loss": 0.3441, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.7987220447284344, | |
| "grad_norm": 0.3568879961967468, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.3302, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.8083067092651754, | |
| "grad_norm": 0.3562428653240204, | |
| "learning_rate": 1.1318413143740436e-07, | |
| "loss": 0.3331, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.817891373801917, | |
| "grad_norm": 0.36870890855789185, | |
| "learning_rate": 1.0162300788382263e-07, | |
| "loss": 0.3351, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.8274760383386583, | |
| "grad_norm": 0.35584965348243713, | |
| "learning_rate": 9.0678523819408e-08, | |
| "loss": 0.3339, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.8370607028753994, | |
| "grad_norm": 0.3848825991153717, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.3383, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.8466453674121404, | |
| "grad_norm": 0.34939953684806824, | |
| "learning_rate": 7.064490740882057e-08, | |
| "loss": 0.3507, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.856230031948882, | |
| "grad_norm": 0.3547082543373108, | |
| "learning_rate": 6.15582970243117e-08, | |
| "loss": 0.3554, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.8658146964856233, | |
| "grad_norm": 0.3536083996295929, | |
| "learning_rate": 5.3093369734816824e-08, | |
| "loss": 0.3632, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.8753993610223643, | |
| "grad_norm": 0.3563961088657379, | |
| "learning_rate": 4.52511911603265e-08, | |
| "loss": 0.3768, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.8849840255591053, | |
| "grad_norm": 0.3636389970779419, | |
| "learning_rate": 3.8032748525179684e-08, | |
| "loss": 0.3407, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.8945686900958467, | |
| "grad_norm": 0.34169384837150574, | |
| "learning_rate": 3.143895053378698e-08, | |
| "loss": 0.3617, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.9041533546325877, | |
| "grad_norm": 0.36579567193984985, | |
| "learning_rate": 2.547062725623828e-08, | |
| "loss": 0.352, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.913738019169329, | |
| "grad_norm": 0.35138216614723206, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.3313, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.92332268370607, | |
| "grad_norm": 0.348357230424881, | |
| "learning_rate": 1.541333133436018e-08, | |
| "loss": 0.3465, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.9329073482428116, | |
| "grad_norm": 0.37804102897644043, | |
| "learning_rate": 1.132562476771959e-08, | |
| "loss": 0.323, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.9424920127795526, | |
| "grad_norm": 0.3628365397453308, | |
| "learning_rate": 7.865924910916977e-09, | |
| "loss": 0.314, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.952076677316294, | |
| "grad_norm": 0.3522026240825653, | |
| "learning_rate": 5.034667293427053e-09, | |
| "loss": 0.3324, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.961661341853035, | |
| "grad_norm": 0.34447744488716125, | |
| "learning_rate": 2.8322083323334417e-09, | |
| "loss": 0.3756, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.9712460063897765, | |
| "grad_norm": 0.3485025465488434, | |
| "learning_rate": 1.2588252874673469e-09, | |
| "loss": 0.3499, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.9808306709265175, | |
| "grad_norm": 0.3362826406955719, | |
| "learning_rate": 3.147162264971471e-10, | |
| "loss": 0.3531, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.9904153354632586, | |
| "grad_norm": 0.3692666292190552, | |
| "learning_rate": 0.0, | |
| "loss": 0.3322, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.9904153354632586, | |
| "step": 312, | |
| "total_flos": 3.4166647452952166e+17, | |
| "train_loss": 0.4181405160671625, | |
| "train_runtime": 9821.1621, | |
| "train_samples_per_second": 3.055, | |
| "train_steps_per_second": 0.032 | |
| } | |
| ], | |
| "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": 3.4166647452952166e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |