| { | |
| "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": 10.041484832763672, | |
| "learning_rate": 3.125e-07, | |
| "loss": 1.2067, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.019169329073482427, | |
| "grad_norm": 9.69267463684082, | |
| "learning_rate": 6.25e-07, | |
| "loss": 1.1824, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.02875399361022364, | |
| "grad_norm": 9.63055419921875, | |
| "learning_rate": 9.375000000000001e-07, | |
| "loss": 1.1705, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.038338658146964855, | |
| "grad_norm": 9.281330108642578, | |
| "learning_rate": 1.25e-06, | |
| "loss": 1.1789, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.04792332268370607, | |
| "grad_norm": 8.621428489685059, | |
| "learning_rate": 1.5625e-06, | |
| "loss": 1.1452, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.05750798722044728, | |
| "grad_norm": 8.266639709472656, | |
| "learning_rate": 1.8750000000000003e-06, | |
| "loss": 1.1485, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0670926517571885, | |
| "grad_norm": 6.1149001121521, | |
| "learning_rate": 2.1875000000000002e-06, | |
| "loss": 1.0678, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.07667731629392971, | |
| "grad_norm": 5.883880615234375, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.0821, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.08626198083067092, | |
| "grad_norm": 3.8237390518188477, | |
| "learning_rate": 2.8125e-06, | |
| "loss": 1.0027, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.09584664536741214, | |
| "grad_norm": 3.546095132827759, | |
| "learning_rate": 3.125e-06, | |
| "loss": 1.0096, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.10543130990415335, | |
| "grad_norm": 3.4178311824798584, | |
| "learning_rate": 3.4375e-06, | |
| "loss": 0.9716, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.11501597444089456, | |
| "grad_norm": 3.771919012069702, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.9502, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.12460063897763578, | |
| "grad_norm": 3.820408344268799, | |
| "learning_rate": 4.0625000000000005e-06, | |
| "loss": 0.9281, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.134185303514377, | |
| "grad_norm": 3.4381327629089355, | |
| "learning_rate": 4.3750000000000005e-06, | |
| "loss": 0.9087, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.14376996805111822, | |
| "grad_norm": 2.9971580505371094, | |
| "learning_rate": 4.6875000000000004e-06, | |
| "loss": 0.903, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.15335463258785942, | |
| "grad_norm": 2.2153100967407227, | |
| "learning_rate": 5e-06, | |
| "loss": 0.8486, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.16293929712460065, | |
| "grad_norm": 2.4878783226013184, | |
| "learning_rate": 5.3125e-06, | |
| "loss": 0.8321, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.17252396166134185, | |
| "grad_norm": 2.3063278198242188, | |
| "learning_rate": 5.625e-06, | |
| "loss": 0.8269, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.18210862619808307, | |
| "grad_norm": 1.750576376914978, | |
| "learning_rate": 5.9375e-06, | |
| "loss": 0.7997, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.19169329073482427, | |
| "grad_norm": 1.4879165887832642, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.8034, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2012779552715655, | |
| "grad_norm": 1.3987113237380981, | |
| "learning_rate": 6.5625e-06, | |
| "loss": 0.7733, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2108626198083067, | |
| "grad_norm": 1.6606000661849976, | |
| "learning_rate": 6.875e-06, | |
| "loss": 0.7515, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.22044728434504793, | |
| "grad_norm": 1.6238853931427002, | |
| "learning_rate": 7.1875e-06, | |
| "loss": 0.7396, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.23003194888178913, | |
| "grad_norm": 1.3396315574645996, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.7438, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.23961661341853036, | |
| "grad_norm": 1.21493661403656, | |
| "learning_rate": 7.8125e-06, | |
| "loss": 0.7533, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.24920127795527156, | |
| "grad_norm": 1.2480034828186035, | |
| "learning_rate": 8.125000000000001e-06, | |
| "loss": 0.7101, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.25878594249201275, | |
| "grad_norm": 1.399794340133667, | |
| "learning_rate": 8.4375e-06, | |
| "loss": 0.7147, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.268370607028754, | |
| "grad_norm": 1.250038981437683, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.7096, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2779552715654952, | |
| "grad_norm": 1.1550706624984741, | |
| "learning_rate": 9.0625e-06, | |
| "loss": 0.6974, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.28753993610223644, | |
| "grad_norm": 1.114206075668335, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 0.6977, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2971246006389776, | |
| "grad_norm": 0.9567247629165649, | |
| "learning_rate": 9.6875e-06, | |
| "loss": 0.6966, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.30670926517571884, | |
| "grad_norm": 1.1043249368667603, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6856, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.31629392971246006, | |
| "grad_norm": 1.1120928525924683, | |
| "learning_rate": 9.999685283773504e-06, | |
| "loss": 0.6977, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3258785942492013, | |
| "grad_norm": 0.8543025851249695, | |
| "learning_rate": 9.998741174712534e-06, | |
| "loss": 0.6773, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.3354632587859425, | |
| "grad_norm": 1.0587968826293945, | |
| "learning_rate": 9.997167791667668e-06, | |
| "loss": 0.6831, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3450479233226837, | |
| "grad_norm": 1.0576103925704956, | |
| "learning_rate": 9.994965332706574e-06, | |
| "loss": 0.6833, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3546325878594249, | |
| "grad_norm": 1.0163214206695557, | |
| "learning_rate": 9.992134075089085e-06, | |
| "loss": 0.6733, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.36421725239616615, | |
| "grad_norm": 0.9684771299362183, | |
| "learning_rate": 9.98867437523228e-06, | |
| "loss": 0.6817, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.3738019169329074, | |
| "grad_norm": 0.8340110778808594, | |
| "learning_rate": 9.984586668665641e-06, | |
| "loss": 0.6656, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.38338658146964855, | |
| "grad_norm": 1.0645322799682617, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.6639, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3929712460063898, | |
| "grad_norm": 0.8854984045028687, | |
| "learning_rate": 9.974529372743762e-06, | |
| "loss": 0.6639, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.402555910543131, | |
| "grad_norm": 1.0209273099899292, | |
| "learning_rate": 9.968561049466214e-06, | |
| "loss": 0.6612, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.41214057507987223, | |
| "grad_norm": 0.9446033239364624, | |
| "learning_rate": 9.961967251474823e-06, | |
| "loss": 0.6525, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4217252396166134, | |
| "grad_norm": 0.8962736129760742, | |
| "learning_rate": 9.954748808839675e-06, | |
| "loss": 0.6451, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.43130990415335463, | |
| "grad_norm": 0.8567093014717102, | |
| "learning_rate": 9.946906630265184e-06, | |
| "loss": 0.6302, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.44089456869009586, | |
| "grad_norm": 1.0150914192199707, | |
| "learning_rate": 9.938441702975689e-06, | |
| "loss": 0.6475, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4504792332268371, | |
| "grad_norm": 0.8952351808547974, | |
| "learning_rate": 9.92935509259118e-06, | |
| "loss": 0.6582, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.46006389776357826, | |
| "grad_norm": 1.0809413194656372, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.6625, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4696485623003195, | |
| "grad_norm": 0.9264349937438965, | |
| "learning_rate": 9.909321476180594e-06, | |
| "loss": 0.6461, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.4792332268370607, | |
| "grad_norm": 0.7585628628730774, | |
| "learning_rate": 9.898376992116179e-06, | |
| "loss": 0.6465, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.48881789137380194, | |
| "grad_norm": 1.0210459232330322, | |
| "learning_rate": 9.886815868562596e-06, | |
| "loss": 0.6519, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4984025559105431, | |
| "grad_norm": 0.8413047194480896, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.6344, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5079872204472844, | |
| "grad_norm": 0.848068118095398, | |
| "learning_rate": 9.861849601988384e-06, | |
| "loss": 0.6535, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5175718849840255, | |
| "grad_norm": 0.8702071905136108, | |
| "learning_rate": 9.848447601883436e-06, | |
| "loss": 0.6366, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5271565495207667, | |
| "grad_norm": 0.8316821455955505, | |
| "learning_rate": 9.834435247725032e-06, | |
| "loss": 0.6328, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.536741214057508, | |
| "grad_norm": 0.9288855791091919, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.6344, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5463258785942492, | |
| "grad_norm": 0.8851747512817383, | |
| "learning_rate": 9.804586609725499e-06, | |
| "loss": 0.6497, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5559105431309904, | |
| "grad_norm": 0.8732710480690002, | |
| "learning_rate": 9.788754083424654e-06, | |
| "loss": 0.6227, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5654952076677316, | |
| "grad_norm": 0.8755611777305603, | |
| "learning_rate": 9.772318717677905e-06, | |
| "loss": 0.6345, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.5750798722044729, | |
| "grad_norm": 1.0519345998764038, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.631, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5846645367412141, | |
| "grad_norm": 0.8587120175361633, | |
| "learning_rate": 9.737647819437645e-06, | |
| "loss": 0.6439, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5942492012779552, | |
| "grad_norm": 0.8681675791740417, | |
| "learning_rate": 9.719416651541839e-06, | |
| "loss": 0.6523, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6038338658146964, | |
| "grad_norm": 1.0971438884735107, | |
| "learning_rate": 9.700591372846096e-06, | |
| "loss": 0.6459, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6134185303514377, | |
| "grad_norm": 0.8674156069755554, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.6153, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.6230031948881789, | |
| "grad_norm": 0.8687777519226074, | |
| "learning_rate": 9.661168036940071e-06, | |
| "loss": 0.6249, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6325878594249201, | |
| "grad_norm": 0.9313914775848389, | |
| "learning_rate": 9.640574942595195e-06, | |
| "loss": 0.6307, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6421725239616614, | |
| "grad_norm": 0.8277938961982727, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.6485, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6517571884984026, | |
| "grad_norm": 0.7965261340141296, | |
| "learning_rate": 9.597638862757255e-06, | |
| "loss": 0.6015, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6613418530351438, | |
| "grad_norm": 0.9038349986076355, | |
| "learning_rate": 9.5753012823366e-06, | |
| "loss": 0.618, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.670926517571885, | |
| "grad_norm": 0.9620576500892639, | |
| "learning_rate": 9.552387733294081e-06, | |
| "loss": 0.6182, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6805111821086262, | |
| "grad_norm": 0.8715839385986328, | |
| "learning_rate": 9.528901100135971e-06, | |
| "loss": 0.6297, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6900958466453674, | |
| "grad_norm": 0.8728576898574829, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.6409, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.6996805111821086, | |
| "grad_norm": 1.0132206678390503, | |
| "learning_rate": 9.480220479843627e-06, | |
| "loss": 0.6418, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7092651757188498, | |
| "grad_norm": 0.905213475227356, | |
| "learning_rate": 9.45503262094184e-06, | |
| "loss": 0.6153, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.7188498402555911, | |
| "grad_norm": 0.7777428030967712, | |
| "learning_rate": 9.4292839336179e-06, | |
| "loss": 0.6039, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7284345047923323, | |
| "grad_norm": 1.0897445678710938, | |
| "learning_rate": 9.40297765928369e-06, | |
| "loss": 0.6089, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.7380191693290735, | |
| "grad_norm": 0.9696489572525024, | |
| "learning_rate": 9.376117109543769e-06, | |
| "loss": 0.6156, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.7476038338658147, | |
| "grad_norm": 0.8268092274665833, | |
| "learning_rate": 9.348705665778479e-06, | |
| "loss": 0.6441, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7571884984025559, | |
| "grad_norm": 1.245557427406311, | |
| "learning_rate": 9.320746778718274e-06, | |
| "loss": 0.6193, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.7667731629392971, | |
| "grad_norm": 0.8154701590538025, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.6334, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7763578274760383, | |
| "grad_norm": 0.9788893461227417, | |
| "learning_rate": 9.263200821770462e-06, | |
| "loss": 0.6302, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7859424920127795, | |
| "grad_norm": 0.9511937499046326, | |
| "learning_rate": 9.233620996141421e-06, | |
| "loss": 0.6146, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7955271565495208, | |
| "grad_norm": 1.0094871520996094, | |
| "learning_rate": 9.203508214822652e-06, | |
| "loss": 0.6115, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.805111821086262, | |
| "grad_norm": 0.9043750762939453, | |
| "learning_rate": 9.172866268606514e-06, | |
| "loss": 0.621, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.8146964856230032, | |
| "grad_norm": 0.8562697172164917, | |
| "learning_rate": 9.141699014900084e-06, | |
| "loss": 0.6176, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8242811501597445, | |
| "grad_norm": 0.8593828082084656, | |
| "learning_rate": 9.110010377239552e-06, | |
| "loss": 0.6378, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.8338658146964856, | |
| "grad_norm": 0.7726905941963196, | |
| "learning_rate": 9.077804344796302e-06, | |
| "loss": 0.6319, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.8434504792332268, | |
| "grad_norm": 0.8027164340019226, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.5939, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.853035143769968, | |
| "grad_norm": 0.798473596572876, | |
| "learning_rate": 9.011856377401891e-06, | |
| "loss": 0.5899, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.8626198083067093, | |
| "grad_norm": 0.8748455047607422, | |
| "learning_rate": 8.978122744408905e-06, | |
| "loss": 0.612, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8722044728434505, | |
| "grad_norm": 0.7628821730613708, | |
| "learning_rate": 8.943888319504456e-06, | |
| "loss": 0.6234, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8817891373801917, | |
| "grad_norm": 0.7844169735908508, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.6128, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8913738019169329, | |
| "grad_norm": 0.8493337035179138, | |
| "learning_rate": 8.873934395068006e-06, | |
| "loss": 0.6141, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.9009584664536742, | |
| "grad_norm": 0.8148529529571533, | |
| "learning_rate": 8.838223701790057e-06, | |
| "loss": 0.5934, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.9105431309904153, | |
| "grad_norm": 0.7671308517456055, | |
| "learning_rate": 8.802029828000157e-06, | |
| "loss": 0.6037, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9201277955271565, | |
| "grad_norm": 0.7300805449485779, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.5961, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.9297124600638977, | |
| "grad_norm": 0.7626590728759766, | |
| "learning_rate": 8.728210824415829e-06, | |
| "loss": 0.604, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.939297124600639, | |
| "grad_norm": 0.8798259496688843, | |
| "learning_rate": 8.690594987436705e-06, | |
| "loss": 0.6075, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.9488817891373802, | |
| "grad_norm": 0.9205968379974365, | |
| "learning_rate": 8.652514554406388e-06, | |
| "loss": 0.6112, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.9584664536741214, | |
| "grad_norm": 0.7867008447647095, | |
| "learning_rate": 8.613974319136959e-06, | |
| "loss": 0.5946, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9680511182108626, | |
| "grad_norm": 0.9884403944015503, | |
| "learning_rate": 8.574979133323378e-06, | |
| "loss": 0.6149, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9776357827476039, | |
| "grad_norm": 0.9288314580917358, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.6076, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.987220447284345, | |
| "grad_norm": 0.9012724757194519, | |
| "learning_rate": 8.495643602586287e-06, | |
| "loss": 0.5918, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9968051118210862, | |
| "grad_norm": 0.8713186979293823, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.6128, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.0063897763578276, | |
| "grad_norm": 0.8621015548706055, | |
| "learning_rate": 8.414547910024035e-06, | |
| "loss": 0.5607, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.0159744408945688, | |
| "grad_norm": 0.766989529132843, | |
| "learning_rate": 8.373352729660373e-06, | |
| "loss": 0.5377, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.0255591054313098, | |
| "grad_norm": 0.8968957662582397, | |
| "learning_rate": 8.331732889760021e-06, | |
| "loss": 0.5361, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.035143769968051, | |
| "grad_norm": 0.7241206765174866, | |
| "learning_rate": 8.289693629698564e-06, | |
| "loss": 0.5248, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.0447284345047922, | |
| "grad_norm": 0.8252156376838684, | |
| "learning_rate": 8.247240241650918e-06, | |
| "loss": 0.5699, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.0543130990415335, | |
| "grad_norm": 0.7180123329162598, | |
| "learning_rate": 8.204378069925121e-06, | |
| "loss": 0.5612, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0638977635782747, | |
| "grad_norm": 0.7721811532974243, | |
| "learning_rate": 8.16111251028955e-06, | |
| "loss": 0.5623, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.073482428115016, | |
| "grad_norm": 0.7802078723907471, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.5223, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.0830670926517572, | |
| "grad_norm": 0.8273614048957825, | |
| "learning_rate": 8.073393063582386e-06, | |
| "loss": 0.5623, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0926517571884984, | |
| "grad_norm": 0.7058457136154175, | |
| "learning_rate": 8.0289502192041e-06, | |
| "loss": 0.5531, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.1022364217252396, | |
| "grad_norm": 0.8331887722015381, | |
| "learning_rate": 7.984126070912519e-06, | |
| "loss": 0.5482, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.1118210862619808, | |
| "grad_norm": 0.7870607376098633, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.5439, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.121405750798722, | |
| "grad_norm": 0.7735496759414673, | |
| "learning_rate": 7.89335648089903e-06, | |
| "loss": 0.5357, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.1309904153354633, | |
| "grad_norm": 0.8142541646957397, | |
| "learning_rate": 7.84742246584226e-06, | |
| "loss": 0.5701, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.1405750798722045, | |
| "grad_norm": 0.7338546514511108, | |
| "learning_rate": 7.801129998764014e-06, | |
| "loss": 0.5478, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.1501597444089458, | |
| "grad_norm": 0.7157395482063293, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.5677, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.159744408945687, | |
| "grad_norm": 0.8188211917877197, | |
| "learning_rate": 7.70749306331863e-06, | |
| "loss": 0.5379, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.1693290734824282, | |
| "grad_norm": 0.7177704572677612, | |
| "learning_rate": 7.660160382576683e-06, | |
| "loss": 0.5382, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.1789137380191694, | |
| "grad_norm": 0.7849264144897461, | |
| "learning_rate": 7.612492823579744e-06, | |
| "loss": 0.5503, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.1884984025559104, | |
| "grad_norm": 0.8151108026504517, | |
| "learning_rate": 7.564496387029532e-06, | |
| "loss": 0.5486, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.1980830670926517, | |
| "grad_norm": 0.8086329102516174, | |
| "learning_rate": 7.516177115029002e-06, | |
| "loss": 0.5354, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.207667731629393, | |
| "grad_norm": 0.6399027705192566, | |
| "learning_rate": 7.467541090321735e-06, | |
| "loss": 0.5435, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.2172523961661341, | |
| "grad_norm": 0.8837975859642029, | |
| "learning_rate": 7.4185944355261996e-06, | |
| "loss": 0.5491, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.2268370607028753, | |
| "grad_norm": 0.6761747002601624, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.5496, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.2364217252396166, | |
| "grad_norm": 0.8938049077987671, | |
| "learning_rate": 7.319793920889171e-06, | |
| "loss": 0.5724, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.2460063897763578, | |
| "grad_norm": 0.6806639432907104, | |
| "learning_rate": 7.269952498697734e-06, | |
| "loss": 0.5535, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.255591054313099, | |
| "grad_norm": 0.6811041831970215, | |
| "learning_rate": 7.219825320152411e-06, | |
| "loss": 0.5419, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.2651757188498403, | |
| "grad_norm": 0.7163997292518616, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.5529, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.2747603833865815, | |
| "grad_norm": 0.7321376204490662, | |
| "learning_rate": 7.118738970516944e-06, | |
| "loss": 0.526, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.2843450479233227, | |
| "grad_norm": 0.699560284614563, | |
| "learning_rate": 7.067792524832604e-06, | |
| "loss": 0.5239, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.293929712460064, | |
| "grad_norm": 0.6956992149353027, | |
| "learning_rate": 7.016585772004026e-06, | |
| "loss": 0.5487, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.3035143769968052, | |
| "grad_norm": 0.684252917766571, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.5307, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.3130990415335464, | |
| "grad_norm": 0.7382246255874634, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.5339, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.3226837060702876, | |
| "grad_norm": 0.693885862827301, | |
| "learning_rate": 6.8614682920097265e-06, | |
| "loss": 0.548, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.3322683706070286, | |
| "grad_norm": 0.6941449642181396, | |
| "learning_rate": 6.809285088483361e-06, | |
| "loss": 0.5611, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.34185303514377, | |
| "grad_norm": 0.7960633635520935, | |
| "learning_rate": 6.7568741204067145e-06, | |
| "loss": 0.5452, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.351437699680511, | |
| "grad_norm": 0.6940522193908691, | |
| "learning_rate": 6.704241985612625e-06, | |
| "loss": 0.5427, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.3610223642172525, | |
| "grad_norm": 0.6586253643035889, | |
| "learning_rate": 6.651395309775837e-06, | |
| "loss": 0.5172, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.3706070287539935, | |
| "grad_norm": 0.6910136342048645, | |
| "learning_rate": 6.598340745578908e-06, | |
| "loss": 0.5403, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.3801916932907348, | |
| "grad_norm": 0.7065456509590149, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.5493, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.389776357827476, | |
| "grad_norm": 0.7206940054893494, | |
| "learning_rate": 6.491634692845781e-06, | |
| "loss": 0.5576, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.3993610223642172, | |
| "grad_norm": 0.6870983839035034, | |
| "learning_rate": 6.437996637160086e-06, | |
| "loss": 0.539, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.4089456869009584, | |
| "grad_norm": 0.7133631706237793, | |
| "learning_rate": 6.384177557124247e-06, | |
| "loss": 0.5378, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.4185303514376997, | |
| "grad_norm": 0.7012754082679749, | |
| "learning_rate": 6.330184227833376e-06, | |
| "loss": 0.5227, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.428115015974441, | |
| "grad_norm": 0.7168461680412292, | |
| "learning_rate": 6.276023446318214e-06, | |
| "loss": 0.5287, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.4376996805111821, | |
| "grad_norm": 0.7431371212005615, | |
| "learning_rate": 6.2217020306894705e-06, | |
| "loss": 0.5746, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4472843450479234, | |
| "grad_norm": 0.8183725476264954, | |
| "learning_rate": 6.1672268192795285e-06, | |
| "loss": 0.5471, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.4568690095846646, | |
| "grad_norm": 0.7414305806159973, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.5305, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.4664536741214058, | |
| "grad_norm": 0.7249237298965454, | |
| "learning_rate": 6.057842458386315e-06, | |
| "loss": 0.539, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.476038338658147, | |
| "grad_norm": 0.7848438620567322, | |
| "learning_rate": 6.002947078916365e-06, | |
| "loss": 0.522, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.4856230031948883, | |
| "grad_norm": 0.6666212677955627, | |
| "learning_rate": 5.947925441958393e-06, | |
| "loss": 0.5366, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.4952076677316293, | |
| "grad_norm": 0.8270851969718933, | |
| "learning_rate": 5.892784473993184e-06, | |
| "loss": 0.5529, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.5047923322683707, | |
| "grad_norm": 0.7611240148544312, | |
| "learning_rate": 5.837531116523683e-06, | |
| "loss": 0.5395, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.5143769968051117, | |
| "grad_norm": 0.691460371017456, | |
| "learning_rate": 5.782172325201155e-06, | |
| "loss": 0.5308, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.5239616613418532, | |
| "grad_norm": 0.7503504157066345, | |
| "learning_rate": 5.726715068949564e-06, | |
| "loss": 0.5589, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.5335463258785942, | |
| "grad_norm": 0.7102341651916504, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.5587, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5431309904153354, | |
| "grad_norm": 0.7421011328697205, | |
| "learning_rate": 5.615533098453215e-06, | |
| "loss": 0.5543, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.5527156549520766, | |
| "grad_norm": 0.7281792163848877, | |
| "learning_rate": 5.559822380516539e-06, | |
| "loss": 0.5305, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.5623003194888179, | |
| "grad_norm": 0.6894298791885376, | |
| "learning_rate": 5.504041188505022e-06, | |
| "loss": 0.5215, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.571884984025559, | |
| "grad_norm": 0.67361980676651, | |
| "learning_rate": 5.448196544517168e-06, | |
| "loss": 0.5459, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.5814696485623003, | |
| "grad_norm": 0.6694360375404358, | |
| "learning_rate": 5.392295478639226e-06, | |
| "loss": 0.5323, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.5910543130990416, | |
| "grad_norm": 0.7042282223701477, | |
| "learning_rate": 5.336345028060199e-06, | |
| "loss": 0.5259, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.6006389776357828, | |
| "grad_norm": 0.6842959523200989, | |
| "learning_rate": 5.2803522361859596e-06, | |
| "loss": 0.5551, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.610223642172524, | |
| "grad_norm": 0.7328696250915527, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.5459, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.619808306709265, | |
| "grad_norm": 0.6389267444610596, | |
| "learning_rate": 5.168267827938971e-06, | |
| "loss": 0.5276, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.6293929712460065, | |
| "grad_norm": 0.7053073644638062, | |
| "learning_rate": 5.112190321479026e-06, | |
| "loss": 0.5485, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6389776357827475, | |
| "grad_norm": 0.7645400166511536, | |
| "learning_rate": 5.05609869177323e-06, | |
| "loss": 0.524, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.648562300319489, | |
| "grad_norm": 0.717299222946167, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5217, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.65814696485623, | |
| "grad_norm": 0.6739091277122498, | |
| "learning_rate": 4.943901308226771e-06, | |
| "loss": 0.5256, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.6677316293929714, | |
| "grad_norm": 0.6906968951225281, | |
| "learning_rate": 4.887809678520976e-06, | |
| "loss": 0.5392, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.6773162939297124, | |
| "grad_norm": 0.6453084349632263, | |
| "learning_rate": 4.831732172061032e-06, | |
| "loss": 0.5464, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.6869009584664538, | |
| "grad_norm": 0.6224555969238281, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.5596, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.6964856230031948, | |
| "grad_norm": 0.7328355312347412, | |
| "learning_rate": 4.719647763814041e-06, | |
| "loss": 0.5299, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.706070287539936, | |
| "grad_norm": 0.6948109865188599, | |
| "learning_rate": 4.663654971939802e-06, | |
| "loss": 0.5504, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.7156549520766773, | |
| "grad_norm": 0.6178452968597412, | |
| "learning_rate": 4.6077045213607765e-06, | |
| "loss": 0.5312, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.7252396166134185, | |
| "grad_norm": 0.651917040348053, | |
| "learning_rate": 4.551803455482833e-06, | |
| "loss": 0.5372, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.7348242811501597, | |
| "grad_norm": 0.672956645488739, | |
| "learning_rate": 4.4959588114949785e-06, | |
| "loss": 0.5285, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.744408945686901, | |
| "grad_norm": 0.6844072937965393, | |
| "learning_rate": 4.4401776194834615e-06, | |
| "loss": 0.5409, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.7539936102236422, | |
| "grad_norm": 0.6573711037635803, | |
| "learning_rate": 4.384466901546786e-06, | |
| "loss": 0.556, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.7635782747603834, | |
| "grad_norm": 0.6577045917510986, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.5417, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.7731629392971247, | |
| "grad_norm": 0.6672411561012268, | |
| "learning_rate": 4.273284931050438e-06, | |
| "loss": 0.5399, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.7827476038338657, | |
| "grad_norm": 0.6693191528320312, | |
| "learning_rate": 4.217827674798845e-06, | |
| "loss": 0.5355, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.792332268370607, | |
| "grad_norm": 0.6733079552650452, | |
| "learning_rate": 4.162468883476319e-06, | |
| "loss": 0.5187, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.8019169329073481, | |
| "grad_norm": 0.7173909544944763, | |
| "learning_rate": 4.107215526006818e-06, | |
| "loss": 0.5334, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.8115015974440896, | |
| "grad_norm": 0.709580659866333, | |
| "learning_rate": 4.052074558041608e-06, | |
| "loss": 0.5396, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.8210862619808306, | |
| "grad_norm": 0.6447334289550781, | |
| "learning_rate": 3.997052921083637e-06, | |
| "loss": 0.5317, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.830670926517572, | |
| "grad_norm": 0.6373368501663208, | |
| "learning_rate": 3.9421575416136866e-06, | |
| "loss": 0.5388, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.840255591054313, | |
| "grad_norm": 0.672348141670227, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.524, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.8498402555910545, | |
| "grad_norm": 0.6385909914970398, | |
| "learning_rate": 3.832773180720475e-06, | |
| "loss": 0.5326, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.8594249201277955, | |
| "grad_norm": 0.6423490047454834, | |
| "learning_rate": 3.778297969310529e-06, | |
| "loss": 0.5274, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.8690095846645367, | |
| "grad_norm": 0.6420339345932007, | |
| "learning_rate": 3.723976553681787e-06, | |
| "loss": 0.5634, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.878594249201278, | |
| "grad_norm": 0.752887487411499, | |
| "learning_rate": 3.669815772166625e-06, | |
| "loss": 0.5351, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.8881789137380192, | |
| "grad_norm": 0.6227625012397766, | |
| "learning_rate": 3.6158224428757538e-06, | |
| "loss": 0.5253, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.8977635782747604, | |
| "grad_norm": 0.5916818976402283, | |
| "learning_rate": 3.562003362839914e-06, | |
| "loss": 0.5327, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.9073482428115016, | |
| "grad_norm": 0.742470920085907, | |
| "learning_rate": 3.50836530715422e-06, | |
| "loss": 0.5583, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.9169329073482428, | |
| "grad_norm": 0.6018843650817871, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.5513, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.926517571884984, | |
| "grad_norm": 0.6124539375305176, | |
| "learning_rate": 3.4016592544210937e-06, | |
| "loss": 0.543, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.9361022364217253, | |
| "grad_norm": 0.6403389573097229, | |
| "learning_rate": 3.3486046902241663e-06, | |
| "loss": 0.531, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.9456869009584663, | |
| "grad_norm": 0.6133631467819214, | |
| "learning_rate": 3.295758014387375e-06, | |
| "loss": 0.5051, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.9552715654952078, | |
| "grad_norm": 0.6370496153831482, | |
| "learning_rate": 3.2431258795932863e-06, | |
| "loss": 0.5461, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.9648562300319488, | |
| "grad_norm": 0.6143704056739807, | |
| "learning_rate": 3.1907149115166403e-06, | |
| "loss": 0.5572, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.9744408945686902, | |
| "grad_norm": 0.6722372770309448, | |
| "learning_rate": 3.1385317079902743e-06, | |
| "loss": 0.5204, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.9840255591054312, | |
| "grad_norm": 0.6286212801933289, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.5419, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.9936102236421727, | |
| "grad_norm": 0.5995206236839294, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.5394, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.0031948881789137, | |
| "grad_norm": 0.6550461053848267, | |
| "learning_rate": 2.9834142279959754e-06, | |
| "loss": 0.5119, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.012779552715655, | |
| "grad_norm": 0.6472194194793701, | |
| "learning_rate": 2.932207475167398e-06, | |
| "loss": 0.4875, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.022364217252396, | |
| "grad_norm": 0.6644472479820251, | |
| "learning_rate": 2.8812610294830568e-06, | |
| "loss": 0.4528, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.0319488817891376, | |
| "grad_norm": 0.5704013705253601, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.4957, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.0415335463258786, | |
| "grad_norm": 0.6226819157600403, | |
| "learning_rate": 2.7801746798475905e-06, | |
| "loss": 0.4965, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.0511182108626196, | |
| "grad_norm": 0.6059399843215942, | |
| "learning_rate": 2.7300475013022666e-06, | |
| "loss": 0.4986, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.060702875399361, | |
| "grad_norm": 0.6234830617904663, | |
| "learning_rate": 2.6802060791108304e-06, | |
| "loss": 0.4885, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.070287539936102, | |
| "grad_norm": 0.6117391586303711, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.4728, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.0798722044728435, | |
| "grad_norm": 0.682826578617096, | |
| "learning_rate": 2.5814055644738013e-06, | |
| "loss": 0.4764, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.0894568690095845, | |
| "grad_norm": 0.6085416078567505, | |
| "learning_rate": 2.532458909678266e-06, | |
| "loss": 0.4823, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.099041533546326, | |
| "grad_norm": 0.6411639451980591, | |
| "learning_rate": 2.483822884971e-06, | |
| "loss": 0.5015, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.108626198083067, | |
| "grad_norm": 0.6138768196105957, | |
| "learning_rate": 2.43550361297047e-06, | |
| "loss": 0.4832, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.1182108626198084, | |
| "grad_norm": 0.6133522987365723, | |
| "learning_rate": 2.387507176420256e-06, | |
| "loss": 0.4701, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.1277955271565494, | |
| "grad_norm": 0.594542920589447, | |
| "learning_rate": 2.339839617423318e-06, | |
| "loss": 0.4765, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.137380191693291, | |
| "grad_norm": 0.6451693773269653, | |
| "learning_rate": 2.2925069366813718e-06, | |
| "loss": 0.4787, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.146964856230032, | |
| "grad_norm": 0.6343329548835754, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.4832, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.1565495207667733, | |
| "grad_norm": 0.6135368347167969, | |
| "learning_rate": 2.1988700012359865e-06, | |
| "loss": 0.4771, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.1661341853035143, | |
| "grad_norm": 0.5679577589035034, | |
| "learning_rate": 2.1525775341577404e-06, | |
| "loss": 0.4973, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.1757188498402558, | |
| "grad_norm": 0.5575957298278809, | |
| "learning_rate": 2.1066435191009717e-06, | |
| "loss": 0.4811, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.1853035143769968, | |
| "grad_norm": 0.5644421577453613, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.4737, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.194888178913738, | |
| "grad_norm": 0.5842664837837219, | |
| "learning_rate": 2.0158739290874822e-06, | |
| "loss": 0.5029, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.2044728434504792, | |
| "grad_norm": 0.5931283831596375, | |
| "learning_rate": 1.971049780795901e-06, | |
| "loss": 0.4768, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.2140575079872207, | |
| "grad_norm": 0.6379035115242004, | |
| "learning_rate": 1.9266069364176144e-06, | |
| "loss": 0.4822, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.2236421725239617, | |
| "grad_norm": 0.6156995892524719, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.4703, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.2332268370607027, | |
| "grad_norm": 0.5634399056434631, | |
| "learning_rate": 1.838887489710452e-06, | |
| "loss": 0.4892, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.242811501597444, | |
| "grad_norm": 0.5468457341194153, | |
| "learning_rate": 1.7956219300748796e-06, | |
| "loss": 0.4735, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.252396166134185, | |
| "grad_norm": 0.566130518913269, | |
| "learning_rate": 1.7527597583490825e-06, | |
| "loss": 0.4898, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.2619808306709266, | |
| "grad_norm": 0.6444130539894104, | |
| "learning_rate": 1.7103063703014372e-06, | |
| "loss": 0.4848, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.2715654952076676, | |
| "grad_norm": 0.5839499235153198, | |
| "learning_rate": 1.6682671102399806e-06, | |
| "loss": 0.4904, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.281150159744409, | |
| "grad_norm": 0.6181128025054932, | |
| "learning_rate": 1.6266472703396286e-06, | |
| "loss": 0.499, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.29073482428115, | |
| "grad_norm": 0.5924781560897827, | |
| "learning_rate": 1.5854520899759656e-06, | |
| "loss": 0.4669, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.3003194888178915, | |
| "grad_norm": 0.548241376876831, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.4737, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.3099041533546325, | |
| "grad_norm": 0.5470595955848694, | |
| "learning_rate": 1.5043563974137132e-06, | |
| "loss": 0.4714, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.319488817891374, | |
| "grad_norm": 0.5618712902069092, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.4788, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.329073482428115, | |
| "grad_norm": 0.6169053912162781, | |
| "learning_rate": 1.4250208666766235e-06, | |
| "loss": 0.4886, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.3386581469648564, | |
| "grad_norm": 0.5996852517127991, | |
| "learning_rate": 1.3860256808630429e-06, | |
| "loss": 0.4687, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.3482428115015974, | |
| "grad_norm": 0.5412963032722473, | |
| "learning_rate": 1.3474854455936126e-06, | |
| "loss": 0.474, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.357827476038339, | |
| "grad_norm": 0.5670292377471924, | |
| "learning_rate": 1.3094050125632973e-06, | |
| "loss": 0.4903, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.36741214057508, | |
| "grad_norm": 0.5631871223449707, | |
| "learning_rate": 1.2717891755841722e-06, | |
| "loss": 0.4833, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.376996805111821, | |
| "grad_norm": 0.5678842663764954, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.4588, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.3865814696485623, | |
| "grad_norm": 0.5656471252441406, | |
| "learning_rate": 1.1979701719998454e-06, | |
| "loss": 0.4725, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.3961661341853033, | |
| "grad_norm": 0.5302391648292542, | |
| "learning_rate": 1.1617762982099446e-06, | |
| "loss": 0.4874, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.405750798722045, | |
| "grad_norm": 0.5217997431755066, | |
| "learning_rate": 1.1260656049319957e-06, | |
| "loss": 0.4856, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.415335463258786, | |
| "grad_norm": 0.5714566707611084, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.4609, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.4249201277955272, | |
| "grad_norm": 0.5429855585098267, | |
| "learning_rate": 1.0561116804955451e-06, | |
| "loss": 0.5047, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.4345047923322682, | |
| "grad_norm": 0.5583947896957397, | |
| "learning_rate": 1.0218772555910955e-06, | |
| "loss": 0.4884, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.4440894568690097, | |
| "grad_norm": 0.5764051675796509, | |
| "learning_rate": 9.881436225981107e-07, | |
| "loss": 0.4758, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.4536741214057507, | |
| "grad_norm": 0.5524209141731262, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.5023, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.463258785942492, | |
| "grad_norm": 0.5488308668136597, | |
| "learning_rate": 9.221956552036992e-07, | |
| "loss": 0.4789, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.472843450479233, | |
| "grad_norm": 0.5414860248565674, | |
| "learning_rate": 8.899896227604509e-07, | |
| "loss": 0.4823, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.4824281150159746, | |
| "grad_norm": 0.5666077136993408, | |
| "learning_rate": 8.58300985099918e-07, | |
| "loss": 0.4615, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.4920127795527156, | |
| "grad_norm": 0.5838495492935181, | |
| "learning_rate": 8.271337313934869e-07, | |
| "loss": 0.464, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.501597444089457, | |
| "grad_norm": 0.5656548738479614, | |
| "learning_rate": 7.964917851773496e-07, | |
| "loss": 0.4767, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.511182108626198, | |
| "grad_norm": 0.5187501907348633, | |
| "learning_rate": 7.663790038585794e-07, | |
| "loss": 0.4777, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.520766773162939, | |
| "grad_norm": 0.5355112552642822, | |
| "learning_rate": 7.367991782295392e-07, | |
| "loss": 0.4855, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.5303514376996805, | |
| "grad_norm": 0.5446317791938782, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.4786, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.539936102236422, | |
| "grad_norm": 0.5624379515647888, | |
| "learning_rate": 6.792532212817271e-07, | |
| "loss": 0.4815, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.549520766773163, | |
| "grad_norm": 0.5310940146446228, | |
| "learning_rate": 6.512943342215234e-07, | |
| "loss": 0.4778, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.559105431309904, | |
| "grad_norm": 0.5407336950302124, | |
| "learning_rate": 6.238828904562316e-07, | |
| "loss": 0.4506, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.5686900958466454, | |
| "grad_norm": 0.5669533014297485, | |
| "learning_rate": 5.9702234071631e-07, | |
| "loss": 0.481, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.5782747603833864, | |
| "grad_norm": 0.5309340357780457, | |
| "learning_rate": 5.707160663821009e-07, | |
| "loss": 0.4851, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.587859424920128, | |
| "grad_norm": 0.5235077738761902, | |
| "learning_rate": 5.449673790581611e-07, | |
| "loss": 0.4478, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.597444089456869, | |
| "grad_norm": 0.552655041217804, | |
| "learning_rate": 5.197795201563744e-07, | |
| "loss": 0.4464, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.6070287539936103, | |
| "grad_norm": 0.5500283241271973, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.489, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.6166134185303513, | |
| "grad_norm": 0.5286134481430054, | |
| "learning_rate": 4.710988998640298e-07, | |
| "loss": 0.4739, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.626198083067093, | |
| "grad_norm": 0.4994671046733856, | |
| "learning_rate": 4.4761226670592074e-07, | |
| "loss": 0.5058, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.635782747603834, | |
| "grad_norm": 0.534156858921051, | |
| "learning_rate": 4.2469871766340096e-07, | |
| "loss": 0.4907, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.6453674121405752, | |
| "grad_norm": 0.5577386021614075, | |
| "learning_rate": 4.0236113724274716e-07, | |
| "loss": 0.4506, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.6549520766773163, | |
| "grad_norm": 0.513982892036438, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.4662, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.6645367412140573, | |
| "grad_norm": 0.5294740796089172, | |
| "learning_rate": 3.5942505740480583e-07, | |
| "loss": 0.4634, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.6741214057507987, | |
| "grad_norm": 0.5422140955924988, | |
| "learning_rate": 3.3883196305992906e-07, | |
| "loss": 0.4673, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.68370607028754, | |
| "grad_norm": 0.5204587578773499, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.4651, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.693290734824281, | |
| "grad_norm": 0.5269270539283752, | |
| "learning_rate": 2.9940862715390483e-07, | |
| "loss": 0.4912, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.702875399361022, | |
| "grad_norm": 0.5300841331481934, | |
| "learning_rate": 2.8058334845816214e-07, | |
| "loss": 0.4722, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.7124600638977636, | |
| "grad_norm": 0.5122048854827881, | |
| "learning_rate": 2.6235218056235633e-07, | |
| "loss": 0.469, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.722044728434505, | |
| "grad_norm": 0.560728907585144, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 0.4698, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.731629392971246, | |
| "grad_norm": 0.5124772787094116, | |
| "learning_rate": 2.276812823220964e-07, | |
| "loss": 0.4446, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.741214057507987, | |
| "grad_norm": 0.5050658583641052, | |
| "learning_rate": 2.1124591657534776e-07, | |
| "loss": 0.4885, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.7507987220447285, | |
| "grad_norm": 0.5437944531440735, | |
| "learning_rate": 1.9541339027450256e-07, | |
| "loss": 0.4689, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.7603833865814695, | |
| "grad_norm": 0.4845607876777649, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.4963, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.769968051118211, | |
| "grad_norm": 0.5433393716812134, | |
| "learning_rate": 1.6556475227496816e-07, | |
| "loss": 0.4848, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.779552715654952, | |
| "grad_norm": 0.5294089317321777, | |
| "learning_rate": 1.5155239811656562e-07, | |
| "loss": 0.4549, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.7891373801916934, | |
| "grad_norm": 0.46811607480049133, | |
| "learning_rate": 1.3815039801161723e-07, | |
| "loss": 0.4949, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.7987220447284344, | |
| "grad_norm": 0.5526321530342102, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.4926, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.8083067092651754, | |
| "grad_norm": 0.4953249394893646, | |
| "learning_rate": 1.1318413143740436e-07, | |
| "loss": 0.5013, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.817891373801917, | |
| "grad_norm": 0.5125919580459595, | |
| "learning_rate": 1.0162300788382263e-07, | |
| "loss": 0.5025, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.8274760383386583, | |
| "grad_norm": 0.5203016996383667, | |
| "learning_rate": 9.0678523819408e-08, | |
| "loss": 0.4653, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.8370607028753994, | |
| "grad_norm": 0.5107254981994629, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.4868, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.8466453674121404, | |
| "grad_norm": 0.5385187268257141, | |
| "learning_rate": 7.064490740882057e-08, | |
| "loss": 0.4773, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.856230031948882, | |
| "grad_norm": 0.5249802470207214, | |
| "learning_rate": 6.15582970243117e-08, | |
| "loss": 0.4753, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.8658146964856233, | |
| "grad_norm": 0.5248013734817505, | |
| "learning_rate": 5.3093369734816824e-08, | |
| "loss": 0.4778, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.8753993610223643, | |
| "grad_norm": 0.4982949495315552, | |
| "learning_rate": 4.52511911603265e-08, | |
| "loss": 0.4856, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.8849840255591053, | |
| "grad_norm": 0.5062425136566162, | |
| "learning_rate": 3.8032748525179684e-08, | |
| "loss": 0.4636, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.8945686900958467, | |
| "grad_norm": 0.524002194404602, | |
| "learning_rate": 3.143895053378698e-08, | |
| "loss": 0.4833, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.9041533546325877, | |
| "grad_norm": 0.5100042223930359, | |
| "learning_rate": 2.547062725623828e-08, | |
| "loss": 0.4598, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.913738019169329, | |
| "grad_norm": 0.498949259519577, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.4842, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.92332268370607, | |
| "grad_norm": 0.5397353172302246, | |
| "learning_rate": 1.541333133436018e-08, | |
| "loss": 0.4549, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.9329073482428116, | |
| "grad_norm": 0.5431365966796875, | |
| "learning_rate": 1.132562476771959e-08, | |
| "loss": 0.4633, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.9424920127795526, | |
| "grad_norm": 0.5180041790008545, | |
| "learning_rate": 7.865924910916977e-09, | |
| "loss": 0.4627, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.952076677316294, | |
| "grad_norm": 0.5015384554862976, | |
| "learning_rate": 5.034667293427053e-09, | |
| "loss": 0.5014, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.961661341853035, | |
| "grad_norm": 0.5198219418525696, | |
| "learning_rate": 2.8322083323334417e-09, | |
| "loss": 0.4546, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.9712460063897765, | |
| "grad_norm": 0.5227149128913879, | |
| "learning_rate": 1.2588252874673469e-09, | |
| "loss": 0.4814, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.9808306709265175, | |
| "grad_norm": 0.518096387386322, | |
| "learning_rate": 3.147162264971471e-10, | |
| "loss": 0.4832, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.9904153354632586, | |
| "grad_norm": 0.5191978812217712, | |
| "learning_rate": 0.0, | |
| "loss": 0.4861, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.9904153354632586, | |
| "step": 312, | |
| "total_flos": 2.172427208694825e+17, | |
| "train_loss": 0.0, | |
| "train_runtime": 6.2928, | |
| "train_samples_per_second": 4767.384, | |
| "train_steps_per_second": 49.581 | |
| } | |
| ], | |
| "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": 2.172427208694825e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |