| { | |
| "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": 8.150485038757324, | |
| "learning_rate": 3.125e-07, | |
| "loss": 1.1263, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.019169329073482427, | |
| "grad_norm": 7.807866096496582, | |
| "learning_rate": 6.25e-07, | |
| "loss": 1.0931, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.02875399361022364, | |
| "grad_norm": 8.187647819519043, | |
| "learning_rate": 9.375000000000001e-07, | |
| "loss": 1.1248, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.038338658146964855, | |
| "grad_norm": 7.566376209259033, | |
| "learning_rate": 1.25e-06, | |
| "loss": 1.0913, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.04792332268370607, | |
| "grad_norm": 7.314204216003418, | |
| "learning_rate": 1.5625e-06, | |
| "loss": 1.0767, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.05750798722044728, | |
| "grad_norm": 7.0795745849609375, | |
| "learning_rate": 1.8750000000000003e-06, | |
| "loss": 1.0705, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0670926517571885, | |
| "grad_norm": 5.6739606857299805, | |
| "learning_rate": 2.1875000000000002e-06, | |
| "loss": 1.0487, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.07667731629392971, | |
| "grad_norm": 5.160563945770264, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.0214, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.08626198083067092, | |
| "grad_norm": 3.2819786071777344, | |
| "learning_rate": 2.8125e-06, | |
| "loss": 0.9842, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.09584664536741214, | |
| "grad_norm": 3.097018003463745, | |
| "learning_rate": 3.125e-06, | |
| "loss": 0.9671, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.10543130990415335, | |
| "grad_norm": 2.9452693462371826, | |
| "learning_rate": 3.4375e-06, | |
| "loss": 0.9597, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.11501597444089456, | |
| "grad_norm": 3.434796094894409, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.9319, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.12460063897763578, | |
| "grad_norm": 4.0431108474731445, | |
| "learning_rate": 4.0625000000000005e-06, | |
| "loss": 0.9196, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.134185303514377, | |
| "grad_norm": 3.556278705596924, | |
| "learning_rate": 4.3750000000000005e-06, | |
| "loss": 0.9054, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.14376996805111822, | |
| "grad_norm": 3.114114761352539, | |
| "learning_rate": 4.6875000000000004e-06, | |
| "loss": 0.8922, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.15335463258785942, | |
| "grad_norm": 2.207472801208496, | |
| "learning_rate": 5e-06, | |
| "loss": 0.8626, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.16293929712460065, | |
| "grad_norm": 2.376880645751953, | |
| "learning_rate": 5.3125e-06, | |
| "loss": 0.8325, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.17252396166134185, | |
| "grad_norm": 2.03688907623291, | |
| "learning_rate": 5.625e-06, | |
| "loss": 0.8278, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.18210862619808307, | |
| "grad_norm": 1.3376281261444092, | |
| "learning_rate": 5.9375e-06, | |
| "loss": 0.801, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.19169329073482427, | |
| "grad_norm": 1.331655502319336, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.7924, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2012779552715655, | |
| "grad_norm": 1.3158297538757324, | |
| "learning_rate": 6.5625e-06, | |
| "loss": 0.7778, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2108626198083067, | |
| "grad_norm": 1.407170295715332, | |
| "learning_rate": 6.875e-06, | |
| "loss": 0.7677, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.22044728434504793, | |
| "grad_norm": 1.3406498432159424, | |
| "learning_rate": 7.1875e-06, | |
| "loss": 0.765, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.23003194888178913, | |
| "grad_norm": 1.0355418920516968, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.7434, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.23961661341853036, | |
| "grad_norm": 1.086203932762146, | |
| "learning_rate": 7.8125e-06, | |
| "loss": 0.7524, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.24920127795527156, | |
| "grad_norm": 1.0376132726669312, | |
| "learning_rate": 8.125000000000001e-06, | |
| "loss": 0.7231, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.25878594249201275, | |
| "grad_norm": 1.0266880989074707, | |
| "learning_rate": 8.4375e-06, | |
| "loss": 0.7251, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.268370607028754, | |
| "grad_norm": 0.8888912796974182, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.722, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2779552715654952, | |
| "grad_norm": 0.8433630466461182, | |
| "learning_rate": 9.0625e-06, | |
| "loss": 0.7107, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.28753993610223644, | |
| "grad_norm": 1.0001300573349, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 0.6975, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2971246006389776, | |
| "grad_norm": 1.0221211910247803, | |
| "learning_rate": 9.6875e-06, | |
| "loss": 0.7019, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.30670926517571884, | |
| "grad_norm": 0.7957472801208496, | |
| "learning_rate": 1e-05, | |
| "loss": 0.7015, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.31629392971246006, | |
| "grad_norm": 0.8188865184783936, | |
| "learning_rate": 9.999685283773504e-06, | |
| "loss": 0.7084, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3258785942492013, | |
| "grad_norm": 0.7626446485519409, | |
| "learning_rate": 9.998741174712534e-06, | |
| "loss": 0.7011, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.3354632587859425, | |
| "grad_norm": 0.8135344386100769, | |
| "learning_rate": 9.997167791667668e-06, | |
| "loss": 0.696, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3450479233226837, | |
| "grad_norm": 0.7661399841308594, | |
| "learning_rate": 9.994965332706574e-06, | |
| "loss": 0.6928, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3546325878594249, | |
| "grad_norm": 0.8579995632171631, | |
| "learning_rate": 9.992134075089085e-06, | |
| "loss": 0.699, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.36421725239616615, | |
| "grad_norm": 0.7115493416786194, | |
| "learning_rate": 9.98867437523228e-06, | |
| "loss": 0.6843, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.3738019169329074, | |
| "grad_norm": 0.7873193025588989, | |
| "learning_rate": 9.984586668665641e-06, | |
| "loss": 0.6867, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.38338658146964855, | |
| "grad_norm": 0.7836501002311707, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.6628, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3929712460063898, | |
| "grad_norm": 0.6418578028678894, | |
| "learning_rate": 9.974529372743762e-06, | |
| "loss": 0.6578, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.402555910543131, | |
| "grad_norm": 0.7128148078918457, | |
| "learning_rate": 9.968561049466214e-06, | |
| "loss": 0.6671, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.41214057507987223, | |
| "grad_norm": 0.6487117409706116, | |
| "learning_rate": 9.961967251474823e-06, | |
| "loss": 0.6648, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4217252396166134, | |
| "grad_norm": 0.6714372634887695, | |
| "learning_rate": 9.954748808839675e-06, | |
| "loss": 0.6499, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.43130990415335463, | |
| "grad_norm": 0.7654340267181396, | |
| "learning_rate": 9.946906630265184e-06, | |
| "loss": 0.6605, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.44089456869009586, | |
| "grad_norm": 0.7319685816764832, | |
| "learning_rate": 9.938441702975689e-06, | |
| "loss": 0.6774, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4504792332268371, | |
| "grad_norm": 0.7591245174407959, | |
| "learning_rate": 9.92935509259118e-06, | |
| "loss": 0.6646, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.46006389776357826, | |
| "grad_norm": 0.7330446839332581, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.6593, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4696485623003195, | |
| "grad_norm": 0.7684568166732788, | |
| "learning_rate": 9.909321476180594e-06, | |
| "loss": 0.6569, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.4792332268370607, | |
| "grad_norm": 0.8853203654289246, | |
| "learning_rate": 9.898376992116179e-06, | |
| "loss": 0.6644, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.48881789137380194, | |
| "grad_norm": 0.6971675157546997, | |
| "learning_rate": 9.886815868562596e-06, | |
| "loss": 0.6522, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4984025559105431, | |
| "grad_norm": 0.7527298331260681, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.6584, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5079872204472844, | |
| "grad_norm": 0.9225770831108093, | |
| "learning_rate": 9.861849601988384e-06, | |
| "loss": 0.6504, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5175718849840255, | |
| "grad_norm": 0.860784113407135, | |
| "learning_rate": 9.848447601883436e-06, | |
| "loss": 0.6642, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5271565495207667, | |
| "grad_norm": 0.7279204726219177, | |
| "learning_rate": 9.834435247725032e-06, | |
| "loss": 0.6349, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.536741214057508, | |
| "grad_norm": 0.8378774523735046, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.625, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5463258785942492, | |
| "grad_norm": 0.6531175971031189, | |
| "learning_rate": 9.804586609725499e-06, | |
| "loss": 0.6583, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5559105431309904, | |
| "grad_norm": 0.7052546739578247, | |
| "learning_rate": 9.788754083424654e-06, | |
| "loss": 0.6352, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5654952076677316, | |
| "grad_norm": 0.7330588698387146, | |
| "learning_rate": 9.772318717677905e-06, | |
| "loss": 0.6464, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.5750798722044729, | |
| "grad_norm": 0.6029200553894043, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.6242, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5846645367412141, | |
| "grad_norm": 0.6845069527626038, | |
| "learning_rate": 9.737647819437645e-06, | |
| "loss": 0.6585, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5942492012779552, | |
| "grad_norm": 0.7698150873184204, | |
| "learning_rate": 9.719416651541839e-06, | |
| "loss": 0.6463, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6038338658146964, | |
| "grad_norm": 0.6855467557907104, | |
| "learning_rate": 9.700591372846096e-06, | |
| "loss": 0.6503, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6134185303514377, | |
| "grad_norm": 0.669455885887146, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.6306, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.6230031948881789, | |
| "grad_norm": 0.7016051411628723, | |
| "learning_rate": 9.661168036940071e-06, | |
| "loss": 0.6462, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6325878594249201, | |
| "grad_norm": 0.6591147184371948, | |
| "learning_rate": 9.640574942595195e-06, | |
| "loss": 0.6363, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6421725239616614, | |
| "grad_norm": 0.6269323825836182, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.6312, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6517571884984026, | |
| "grad_norm": 0.6276299953460693, | |
| "learning_rate": 9.597638862757255e-06, | |
| "loss": 0.6485, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6613418530351438, | |
| "grad_norm": 0.6555031538009644, | |
| "learning_rate": 9.5753012823366e-06, | |
| "loss": 0.6271, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.670926517571885, | |
| "grad_norm": 0.617702841758728, | |
| "learning_rate": 9.552387733294081e-06, | |
| "loss": 0.6405, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6805111821086262, | |
| "grad_norm": 0.5773834586143494, | |
| "learning_rate": 9.528901100135971e-06, | |
| "loss": 0.6462, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6900958466453674, | |
| "grad_norm": 0.5768316388130188, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.6494, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.6996805111821086, | |
| "grad_norm": 0.6002576947212219, | |
| "learning_rate": 9.480220479843627e-06, | |
| "loss": 0.643, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7092651757188498, | |
| "grad_norm": 0.6929178833961487, | |
| "learning_rate": 9.45503262094184e-06, | |
| "loss": 0.6449, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.7188498402555911, | |
| "grad_norm": 0.633795440196991, | |
| "learning_rate": 9.4292839336179e-06, | |
| "loss": 0.6438, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7284345047923323, | |
| "grad_norm": 0.6836037635803223, | |
| "learning_rate": 9.40297765928369e-06, | |
| "loss": 0.6277, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.7380191693290735, | |
| "grad_norm": 0.5767949223518372, | |
| "learning_rate": 9.376117109543769e-06, | |
| "loss": 0.6346, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.7476038338658147, | |
| "grad_norm": 0.6540945172309875, | |
| "learning_rate": 9.348705665778479e-06, | |
| "loss": 0.6398, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7571884984025559, | |
| "grad_norm": 0.6413583159446716, | |
| "learning_rate": 9.320746778718274e-06, | |
| "loss": 0.6214, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.7667731629392971, | |
| "grad_norm": 0.6472015380859375, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.6359, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7763578274760383, | |
| "grad_norm": 0.5751689672470093, | |
| "learning_rate": 9.263200821770462e-06, | |
| "loss": 0.6387, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7859424920127795, | |
| "grad_norm": 0.8359308242797852, | |
| "learning_rate": 9.233620996141421e-06, | |
| "loss": 0.6336, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7955271565495208, | |
| "grad_norm": 0.7323636412620544, | |
| "learning_rate": 9.203508214822652e-06, | |
| "loss": 0.6361, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.805111821086262, | |
| "grad_norm": 0.6433335542678833, | |
| "learning_rate": 9.172866268606514e-06, | |
| "loss": 0.6283, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.8146964856230032, | |
| "grad_norm": 0.7690201997756958, | |
| "learning_rate": 9.141699014900084e-06, | |
| "loss": 0.6334, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8242811501597445, | |
| "grad_norm": 0.7022820115089417, | |
| "learning_rate": 9.110010377239552e-06, | |
| "loss": 0.6166, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.8338658146964856, | |
| "grad_norm": 0.7924836277961731, | |
| "learning_rate": 9.077804344796302e-06, | |
| "loss": 0.6454, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.8434504792332268, | |
| "grad_norm": 0.7118843197822571, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.6384, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.853035143769968, | |
| "grad_norm": 0.717670202255249, | |
| "learning_rate": 9.011856377401891e-06, | |
| "loss": 0.6347, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.8626198083067093, | |
| "grad_norm": 0.7034146189689636, | |
| "learning_rate": 8.978122744408905e-06, | |
| "loss": 0.6206, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8722044728434505, | |
| "grad_norm": 0.674736499786377, | |
| "learning_rate": 8.943888319504456e-06, | |
| "loss": 0.6217, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8817891373801917, | |
| "grad_norm": 0.6070693731307983, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.6326, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8913738019169329, | |
| "grad_norm": 0.6934514045715332, | |
| "learning_rate": 8.873934395068006e-06, | |
| "loss": 0.6109, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.9009584664536742, | |
| "grad_norm": 0.6248501539230347, | |
| "learning_rate": 8.838223701790057e-06, | |
| "loss": 0.6316, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.9105431309904153, | |
| "grad_norm": 0.5615552663803101, | |
| "learning_rate": 8.802029828000157e-06, | |
| "loss": 0.6261, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9201277955271565, | |
| "grad_norm": 0.598068118095398, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.6406, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.9297124600638977, | |
| "grad_norm": 0.8680521249771118, | |
| "learning_rate": 8.728210824415829e-06, | |
| "loss": 0.6357, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.939297124600639, | |
| "grad_norm": 0.5738173127174377, | |
| "learning_rate": 8.690594987436705e-06, | |
| "loss": 0.6032, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.9488817891373802, | |
| "grad_norm": 0.73792564868927, | |
| "learning_rate": 8.652514554406388e-06, | |
| "loss": 0.5996, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.9584664536741214, | |
| "grad_norm": 0.7462478876113892, | |
| "learning_rate": 8.613974319136959e-06, | |
| "loss": 0.6229, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9680511182108626, | |
| "grad_norm": 0.6320640444755554, | |
| "learning_rate": 8.574979133323378e-06, | |
| "loss": 0.6317, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9776357827476039, | |
| "grad_norm": 0.7389762997627258, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.631, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.987220447284345, | |
| "grad_norm": 0.6447233557701111, | |
| "learning_rate": 8.495643602586287e-06, | |
| "loss": 0.6205, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9968051118210862, | |
| "grad_norm": 0.6227617859840393, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.614, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.0063897763578276, | |
| "grad_norm": 0.7328119277954102, | |
| "learning_rate": 8.414547910024035e-06, | |
| "loss": 0.5928, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.0159744408945688, | |
| "grad_norm": 0.6177286505699158, | |
| "learning_rate": 8.373352729660373e-06, | |
| "loss": 0.587, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.0255591054313098, | |
| "grad_norm": 0.6559479236602783, | |
| "learning_rate": 8.331732889760021e-06, | |
| "loss": 0.5827, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.035143769968051, | |
| "grad_norm": 0.718102753162384, | |
| "learning_rate": 8.289693629698564e-06, | |
| "loss": 0.5899, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.0447284345047922, | |
| "grad_norm": 0.6655651926994324, | |
| "learning_rate": 8.247240241650918e-06, | |
| "loss": 0.588, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.0543130990415335, | |
| "grad_norm": 0.7050108909606934, | |
| "learning_rate": 8.204378069925121e-06, | |
| "loss": 0.5913, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0638977635782747, | |
| "grad_norm": 0.715863049030304, | |
| "learning_rate": 8.16111251028955e-06, | |
| "loss": 0.5808, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.073482428115016, | |
| "grad_norm": 0.6538060903549194, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.5791, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.0830670926517572, | |
| "grad_norm": 0.6200713515281677, | |
| "learning_rate": 8.073393063582386e-06, | |
| "loss": 0.5707, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0926517571884984, | |
| "grad_norm": 0.6361571550369263, | |
| "learning_rate": 8.0289502192041e-06, | |
| "loss": 0.5739, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.1022364217252396, | |
| "grad_norm": 0.6072981953620911, | |
| "learning_rate": 7.984126070912519e-06, | |
| "loss": 0.5665, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.1118210862619808, | |
| "grad_norm": 0.5768544673919678, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.5648, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.121405750798722, | |
| "grad_norm": 0.6660125255584717, | |
| "learning_rate": 7.89335648089903e-06, | |
| "loss": 0.5893, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.1309904153354633, | |
| "grad_norm": 0.6252293586730957, | |
| "learning_rate": 7.84742246584226e-06, | |
| "loss": 0.5847, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.1405750798722045, | |
| "grad_norm": 0.5960403680801392, | |
| "learning_rate": 7.801129998764014e-06, | |
| "loss": 0.5734, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.1501597444089458, | |
| "grad_norm": 0.7005594372749329, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.5751, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.159744408945687, | |
| "grad_norm": 0.6234583258628845, | |
| "learning_rate": 7.70749306331863e-06, | |
| "loss": 0.5652, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.1693290734824282, | |
| "grad_norm": 0.5803803205490112, | |
| "learning_rate": 7.660160382576683e-06, | |
| "loss": 0.5717, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.1789137380191694, | |
| "grad_norm": 0.6783732175827026, | |
| "learning_rate": 7.612492823579744e-06, | |
| "loss": 0.5789, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.1884984025559104, | |
| "grad_norm": 0.5564832091331482, | |
| "learning_rate": 7.564496387029532e-06, | |
| "loss": 0.5678, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.1980830670926517, | |
| "grad_norm": 0.675945520401001, | |
| "learning_rate": 7.516177115029002e-06, | |
| "loss": 0.5741, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.207667731629393, | |
| "grad_norm": 0.6526042819023132, | |
| "learning_rate": 7.467541090321735e-06, | |
| "loss": 0.5587, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.2172523961661341, | |
| "grad_norm": 0.6318403482437134, | |
| "learning_rate": 7.4185944355261996e-06, | |
| "loss": 0.5796, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.2268370607028753, | |
| "grad_norm": 0.6410267949104309, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.5808, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.2364217252396166, | |
| "grad_norm": 0.6449666023254395, | |
| "learning_rate": 7.319793920889171e-06, | |
| "loss": 0.5783, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.2460063897763578, | |
| "grad_norm": 0.6435701847076416, | |
| "learning_rate": 7.269952498697734e-06, | |
| "loss": 0.5577, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.255591054313099, | |
| "grad_norm": 0.7610958814620972, | |
| "learning_rate": 7.219825320152411e-06, | |
| "loss": 0.5508, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.2651757188498403, | |
| "grad_norm": 0.5984666347503662, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.5682, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.2747603833865815, | |
| "grad_norm": 0.6728739738464355, | |
| "learning_rate": 7.118738970516944e-06, | |
| "loss": 0.5719, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.2843450479233227, | |
| "grad_norm": 0.544328510761261, | |
| "learning_rate": 7.067792524832604e-06, | |
| "loss": 0.566, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.293929712460064, | |
| "grad_norm": 0.6412566900253296, | |
| "learning_rate": 7.016585772004026e-06, | |
| "loss": 0.5792, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.3035143769968052, | |
| "grad_norm": 0.6340923309326172, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.579, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.3130990415335464, | |
| "grad_norm": 0.634636640548706, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.5663, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.3226837060702876, | |
| "grad_norm": 0.6472147703170776, | |
| "learning_rate": 6.8614682920097265e-06, | |
| "loss": 0.581, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.3322683706070286, | |
| "grad_norm": 0.5313640832901001, | |
| "learning_rate": 6.809285088483361e-06, | |
| "loss": 0.5713, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.34185303514377, | |
| "grad_norm": 0.6307949423789978, | |
| "learning_rate": 6.7568741204067145e-06, | |
| "loss": 0.5703, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.351437699680511, | |
| "grad_norm": 0.5627026557922363, | |
| "learning_rate": 6.704241985612625e-06, | |
| "loss": 0.5561, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.3610223642172525, | |
| "grad_norm": 0.6823342442512512, | |
| "learning_rate": 6.651395309775837e-06, | |
| "loss": 0.5614, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.3706070287539935, | |
| "grad_norm": 0.5905282497406006, | |
| "learning_rate": 6.598340745578908e-06, | |
| "loss": 0.5766, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.3801916932907348, | |
| "grad_norm": 0.7758175134658813, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.5611, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.389776357827476, | |
| "grad_norm": 0.518359363079071, | |
| "learning_rate": 6.491634692845781e-06, | |
| "loss": 0.5691, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.3993610223642172, | |
| "grad_norm": 0.7256923317909241, | |
| "learning_rate": 6.437996637160086e-06, | |
| "loss": 0.5694, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.4089456869009584, | |
| "grad_norm": 0.6003644466400146, | |
| "learning_rate": 6.384177557124247e-06, | |
| "loss": 0.5592, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.4185303514376997, | |
| "grad_norm": 0.6105974912643433, | |
| "learning_rate": 6.330184227833376e-06, | |
| "loss": 0.5804, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.428115015974441, | |
| "grad_norm": 0.6344504356384277, | |
| "learning_rate": 6.276023446318214e-06, | |
| "loss": 0.5581, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.4376996805111821, | |
| "grad_norm": 0.6138957142829895, | |
| "learning_rate": 6.2217020306894705e-06, | |
| "loss": 0.5688, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4472843450479234, | |
| "grad_norm": 0.547579824924469, | |
| "learning_rate": 6.1672268192795285e-06, | |
| "loss": 0.5635, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.4568690095846646, | |
| "grad_norm": 0.5572775602340698, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.577, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.4664536741214058, | |
| "grad_norm": 0.5766530632972717, | |
| "learning_rate": 6.057842458386315e-06, | |
| "loss": 0.5775, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.476038338658147, | |
| "grad_norm": 0.5293987989425659, | |
| "learning_rate": 6.002947078916365e-06, | |
| "loss": 0.5714, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.4856230031948883, | |
| "grad_norm": 0.5267698764801025, | |
| "learning_rate": 5.947925441958393e-06, | |
| "loss": 0.5621, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.4952076677316293, | |
| "grad_norm": 0.5576557517051697, | |
| "learning_rate": 5.892784473993184e-06, | |
| "loss": 0.566, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.5047923322683707, | |
| "grad_norm": 0.5801796317100525, | |
| "learning_rate": 5.837531116523683e-06, | |
| "loss": 0.5796, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.5143769968051117, | |
| "grad_norm": 0.546893298625946, | |
| "learning_rate": 5.782172325201155e-06, | |
| "loss": 0.5699, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.5239616613418532, | |
| "grad_norm": 0.5638102293014526, | |
| "learning_rate": 5.726715068949564e-06, | |
| "loss": 0.5639, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.5335463258785942, | |
| "grad_norm": 0.6294999718666077, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.5747, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5431309904153354, | |
| "grad_norm": 0.5858036875724792, | |
| "learning_rate": 5.615533098453215e-06, | |
| "loss": 0.5574, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.5527156549520766, | |
| "grad_norm": 0.6968753933906555, | |
| "learning_rate": 5.559822380516539e-06, | |
| "loss": 0.5602, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.5623003194888179, | |
| "grad_norm": 0.5758054852485657, | |
| "learning_rate": 5.504041188505022e-06, | |
| "loss": 0.5751, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.571884984025559, | |
| "grad_norm": 0.5420729517936707, | |
| "learning_rate": 5.448196544517168e-06, | |
| "loss": 0.5731, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.5814696485623003, | |
| "grad_norm": 0.6788923144340515, | |
| "learning_rate": 5.392295478639226e-06, | |
| "loss": 0.5751, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.5910543130990416, | |
| "grad_norm": 0.6043679714202881, | |
| "learning_rate": 5.336345028060199e-06, | |
| "loss": 0.5701, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.6006389776357828, | |
| "grad_norm": 0.528847873210907, | |
| "learning_rate": 5.2803522361859596e-06, | |
| "loss": 0.5745, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.610223642172524, | |
| "grad_norm": 0.5885607600212097, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.5622, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.619808306709265, | |
| "grad_norm": 0.5792953372001648, | |
| "learning_rate": 5.168267827938971e-06, | |
| "loss": 0.5571, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.6293929712460065, | |
| "grad_norm": 0.5489832162857056, | |
| "learning_rate": 5.112190321479026e-06, | |
| "loss": 0.5696, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6389776357827475, | |
| "grad_norm": 0.6276931166648865, | |
| "learning_rate": 5.05609869177323e-06, | |
| "loss": 0.6019, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.648562300319489, | |
| "grad_norm": 0.5779969692230225, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5767, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.65814696485623, | |
| "grad_norm": 0.5757343173027039, | |
| "learning_rate": 4.943901308226771e-06, | |
| "loss": 0.5611, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.6677316293929714, | |
| "grad_norm": 0.585114598274231, | |
| "learning_rate": 4.887809678520976e-06, | |
| "loss": 0.5498, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.6773162939297124, | |
| "grad_norm": 0.5320142507553101, | |
| "learning_rate": 4.831732172061032e-06, | |
| "loss": 0.5793, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.6869009584664538, | |
| "grad_norm": 0.6053058505058289, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.5616, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.6964856230031948, | |
| "grad_norm": 0.663553774356842, | |
| "learning_rate": 4.719647763814041e-06, | |
| "loss": 0.5763, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.706070287539936, | |
| "grad_norm": 0.5003867149353027, | |
| "learning_rate": 4.663654971939802e-06, | |
| "loss": 0.579, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.7156549520766773, | |
| "grad_norm": 0.6177144646644592, | |
| "learning_rate": 4.6077045213607765e-06, | |
| "loss": 0.5611, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.7252396166134185, | |
| "grad_norm": 0.5344464182853699, | |
| "learning_rate": 4.551803455482833e-06, | |
| "loss": 0.5778, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.7348242811501597, | |
| "grad_norm": 0.5539312362670898, | |
| "learning_rate": 4.4959588114949785e-06, | |
| "loss": 0.5686, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.744408945686901, | |
| "grad_norm": 0.5916006565093994, | |
| "learning_rate": 4.4401776194834615e-06, | |
| "loss": 0.5692, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.7539936102236422, | |
| "grad_norm": 0.6805198192596436, | |
| "learning_rate": 4.384466901546786e-06, | |
| "loss": 0.5762, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.7635782747603834, | |
| "grad_norm": 0.5966302156448364, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.5489, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.7731629392971247, | |
| "grad_norm": 0.6216447949409485, | |
| "learning_rate": 4.273284931050438e-06, | |
| "loss": 0.5634, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.7827476038338657, | |
| "grad_norm": 0.7536059617996216, | |
| "learning_rate": 4.217827674798845e-06, | |
| "loss": 0.566, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.792332268370607, | |
| "grad_norm": 0.5549724698066711, | |
| "learning_rate": 4.162468883476319e-06, | |
| "loss": 0.5837, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.8019169329073481, | |
| "grad_norm": 0.599871814250946, | |
| "learning_rate": 4.107215526006818e-06, | |
| "loss": 0.5646, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.8115015974440896, | |
| "grad_norm": 0.6804283261299133, | |
| "learning_rate": 4.052074558041608e-06, | |
| "loss": 0.5644, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.8210862619808306, | |
| "grad_norm": 0.5554932951927185, | |
| "learning_rate": 3.997052921083637e-06, | |
| "loss": 0.5576, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.830670926517572, | |
| "grad_norm": 0.5054816007614136, | |
| "learning_rate": 3.9421575416136866e-06, | |
| "loss": 0.5701, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.840255591054313, | |
| "grad_norm": 0.5952098965644836, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.5578, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.8498402555910545, | |
| "grad_norm": 0.5662181377410889, | |
| "learning_rate": 3.832773180720475e-06, | |
| "loss": 0.5652, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.8594249201277955, | |
| "grad_norm": 0.5168705582618713, | |
| "learning_rate": 3.778297969310529e-06, | |
| "loss": 0.5649, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.8690095846645367, | |
| "grad_norm": 0.5198647975921631, | |
| "learning_rate": 3.723976553681787e-06, | |
| "loss": 0.5648, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.878594249201278, | |
| "grad_norm": 0.5528584122657776, | |
| "learning_rate": 3.669815772166625e-06, | |
| "loss": 0.5738, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.8881789137380192, | |
| "grad_norm": 0.6017258167266846, | |
| "learning_rate": 3.6158224428757538e-06, | |
| "loss": 0.5735, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.8977635782747604, | |
| "grad_norm": 0.5080370306968689, | |
| "learning_rate": 3.562003362839914e-06, | |
| "loss": 0.5731, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.9073482428115016, | |
| "grad_norm": 0.5272023677825928, | |
| "learning_rate": 3.50836530715422e-06, | |
| "loss": 0.5617, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.9169329073482428, | |
| "grad_norm": 0.5592118501663208, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.5692, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.926517571884984, | |
| "grad_norm": 0.4761984646320343, | |
| "learning_rate": 3.4016592544210937e-06, | |
| "loss": 0.5784, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.9361022364217253, | |
| "grad_norm": 0.4980798661708832, | |
| "learning_rate": 3.3486046902241663e-06, | |
| "loss": 0.5664, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.9456869009584663, | |
| "grad_norm": 0.49683818221092224, | |
| "learning_rate": 3.295758014387375e-06, | |
| "loss": 0.5673, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.9552715654952078, | |
| "grad_norm": 0.49688446521759033, | |
| "learning_rate": 3.2431258795932863e-06, | |
| "loss": 0.5604, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.9648562300319488, | |
| "grad_norm": 0.5208616256713867, | |
| "learning_rate": 3.1907149115166403e-06, | |
| "loss": 0.5622, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.9744408945686902, | |
| "grad_norm": 0.5107452869415283, | |
| "learning_rate": 3.1385317079902743e-06, | |
| "loss": 0.5767, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.9840255591054312, | |
| "grad_norm": 0.5208856463432312, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.5667, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.9936102236421727, | |
| "grad_norm": 0.48777708411216736, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.5576, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.0031948881789137, | |
| "grad_norm": 0.5471110939979553, | |
| "learning_rate": 2.9834142279959754e-06, | |
| "loss": 0.5433, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.012779552715655, | |
| "grad_norm": 0.5684826374053955, | |
| "learning_rate": 2.932207475167398e-06, | |
| "loss": 0.5361, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.022364217252396, | |
| "grad_norm": 0.5112109184265137, | |
| "learning_rate": 2.8812610294830568e-06, | |
| "loss": 0.5192, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.0319488817891376, | |
| "grad_norm": 0.494600385427475, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.5419, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.0415335463258786, | |
| "grad_norm": 0.4980970323085785, | |
| "learning_rate": 2.7801746798475905e-06, | |
| "loss": 0.5036, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.0511182108626196, | |
| "grad_norm": 0.5328761339187622, | |
| "learning_rate": 2.7300475013022666e-06, | |
| "loss": 0.5332, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.060702875399361, | |
| "grad_norm": 0.5595293045043945, | |
| "learning_rate": 2.6802060791108304e-06, | |
| "loss": 0.5387, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.070287539936102, | |
| "grad_norm": 0.4666793644428253, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.5017, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.0798722044728435, | |
| "grad_norm": 0.4883419871330261, | |
| "learning_rate": 2.5814055644738013e-06, | |
| "loss": 0.5279, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.0894568690095845, | |
| "grad_norm": 0.5053743720054626, | |
| "learning_rate": 2.532458909678266e-06, | |
| "loss": 0.5164, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.099041533546326, | |
| "grad_norm": 0.4824915826320648, | |
| "learning_rate": 2.483822884971e-06, | |
| "loss": 0.5454, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.108626198083067, | |
| "grad_norm": 0.5914897322654724, | |
| "learning_rate": 2.43550361297047e-06, | |
| "loss": 0.5244, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.1182108626198084, | |
| "grad_norm": 0.5082386136054993, | |
| "learning_rate": 2.387507176420256e-06, | |
| "loss": 0.5386, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.1277955271565494, | |
| "grad_norm": 0.4908734858036041, | |
| "learning_rate": 2.339839617423318e-06, | |
| "loss": 0.5231, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.137380191693291, | |
| "grad_norm": 0.5613354444503784, | |
| "learning_rate": 2.2925069366813718e-06, | |
| "loss": 0.5303, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.146964856230032, | |
| "grad_norm": 0.47516801953315735, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.5139, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.1565495207667733, | |
| "grad_norm": 0.4658454656600952, | |
| "learning_rate": 2.1988700012359865e-06, | |
| "loss": 0.5277, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.1661341853035143, | |
| "grad_norm": 0.4728820323944092, | |
| "learning_rate": 2.1525775341577404e-06, | |
| "loss": 0.53, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.1757188498402558, | |
| "grad_norm": 0.6010584235191345, | |
| "learning_rate": 2.1066435191009717e-06, | |
| "loss": 0.5287, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.1853035143769968, | |
| "grad_norm": 0.5559404492378235, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.5112, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.194888178913738, | |
| "grad_norm": 0.48023709654808044, | |
| "learning_rate": 2.0158739290874822e-06, | |
| "loss": 0.5221, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.2044728434504792, | |
| "grad_norm": 0.44798821210861206, | |
| "learning_rate": 1.971049780795901e-06, | |
| "loss": 0.5184, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.2140575079872207, | |
| "grad_norm": 0.4652976393699646, | |
| "learning_rate": 1.9266069364176144e-06, | |
| "loss": 0.533, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.2236421725239617, | |
| "grad_norm": 0.47819817066192627, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.5029, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.2332268370607027, | |
| "grad_norm": 0.4362553656101227, | |
| "learning_rate": 1.838887489710452e-06, | |
| "loss": 0.525, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.242811501597444, | |
| "grad_norm": 0.5460666418075562, | |
| "learning_rate": 1.7956219300748796e-06, | |
| "loss": 0.5082, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.252396166134185, | |
| "grad_norm": 0.516977846622467, | |
| "learning_rate": 1.7527597583490825e-06, | |
| "loss": 0.5324, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.2619808306709266, | |
| "grad_norm": 0.4478202760219574, | |
| "learning_rate": 1.7103063703014372e-06, | |
| "loss": 0.534, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.2715654952076676, | |
| "grad_norm": 0.4357537031173706, | |
| "learning_rate": 1.6682671102399806e-06, | |
| "loss": 0.5375, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.281150159744409, | |
| "grad_norm": 0.4697173535823822, | |
| "learning_rate": 1.6266472703396286e-06, | |
| "loss": 0.539, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.29073482428115, | |
| "grad_norm": 0.4569830298423767, | |
| "learning_rate": 1.5854520899759656e-06, | |
| "loss": 0.5456, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.3003194888178915, | |
| "grad_norm": 0.4430941641330719, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.5117, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.3099041533546325, | |
| "grad_norm": 0.4393070340156555, | |
| "learning_rate": 1.5043563974137132e-06, | |
| "loss": 0.5448, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.319488817891374, | |
| "grad_norm": 0.43000754714012146, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.5259, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.329073482428115, | |
| "grad_norm": 0.4374457001686096, | |
| "learning_rate": 1.4250208666766235e-06, | |
| "loss": 0.5177, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.3386581469648564, | |
| "grad_norm": 0.46506327390670776, | |
| "learning_rate": 1.3860256808630429e-06, | |
| "loss": 0.5201, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.3482428115015974, | |
| "grad_norm": 0.4384128749370575, | |
| "learning_rate": 1.3474854455936126e-06, | |
| "loss": 0.5359, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.357827476038339, | |
| "grad_norm": 0.4446409344673157, | |
| "learning_rate": 1.3094050125632973e-06, | |
| "loss": 0.5185, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.36741214057508, | |
| "grad_norm": 0.4300197660923004, | |
| "learning_rate": 1.2717891755841722e-06, | |
| "loss": 0.5141, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.376996805111821, | |
| "grad_norm": 0.44469180703163147, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.5053, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.3865814696485623, | |
| "grad_norm": 0.4117085933685303, | |
| "learning_rate": 1.1979701719998454e-06, | |
| "loss": 0.5369, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.3961661341853033, | |
| "grad_norm": 0.431255042552948, | |
| "learning_rate": 1.1617762982099446e-06, | |
| "loss": 0.5194, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.405750798722045, | |
| "grad_norm": 0.4269045889377594, | |
| "learning_rate": 1.1260656049319957e-06, | |
| "loss": 0.5162, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.415335463258786, | |
| "grad_norm": 0.4201407730579376, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.5252, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.4249201277955272, | |
| "grad_norm": 0.42810115218162537, | |
| "learning_rate": 1.0561116804955451e-06, | |
| "loss": 0.5318, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.4345047923322682, | |
| "grad_norm": 0.4129791855812073, | |
| "learning_rate": 1.0218772555910955e-06, | |
| "loss": 0.5145, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.4440894568690097, | |
| "grad_norm": 0.39486438035964966, | |
| "learning_rate": 9.881436225981107e-07, | |
| "loss": 0.5252, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.4536741214057507, | |
| "grad_norm": 0.4195745289325714, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.5261, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.463258785942492, | |
| "grad_norm": 0.3990418314933777, | |
| "learning_rate": 9.221956552036992e-07, | |
| "loss": 0.5209, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.472843450479233, | |
| "grad_norm": 0.44864341616630554, | |
| "learning_rate": 8.899896227604509e-07, | |
| "loss": 0.516, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.4824281150159746, | |
| "grad_norm": 0.4024929404258728, | |
| "learning_rate": 8.58300985099918e-07, | |
| "loss": 0.5253, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.4920127795527156, | |
| "grad_norm": 0.4194217026233673, | |
| "learning_rate": 8.271337313934869e-07, | |
| "loss": 0.5119, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.501597444089457, | |
| "grad_norm": 0.4190581440925598, | |
| "learning_rate": 7.964917851773496e-07, | |
| "loss": 0.526, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.511182108626198, | |
| "grad_norm": 0.4602578580379486, | |
| "learning_rate": 7.663790038585794e-07, | |
| "loss": 0.512, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.520766773162939, | |
| "grad_norm": 0.38818320631980896, | |
| "learning_rate": 7.367991782295392e-07, | |
| "loss": 0.5292, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.5303514376996805, | |
| "grad_norm": 0.40392065048217773, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.5347, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.539936102236422, | |
| "grad_norm": 0.41519656777381897, | |
| "learning_rate": 6.792532212817271e-07, | |
| "loss": 0.5176, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.549520766773163, | |
| "grad_norm": 0.44370123744010925, | |
| "learning_rate": 6.512943342215234e-07, | |
| "loss": 0.5404, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.559105431309904, | |
| "grad_norm": 0.4083189070224762, | |
| "learning_rate": 6.238828904562316e-07, | |
| "loss": 0.5133, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.5686900958466454, | |
| "grad_norm": 0.4141216576099396, | |
| "learning_rate": 5.9702234071631e-07, | |
| "loss": 0.5377, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.5782747603833864, | |
| "grad_norm": 0.45661696791648865, | |
| "learning_rate": 5.707160663821009e-07, | |
| "loss": 0.5087, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.587859424920128, | |
| "grad_norm": 0.426259845495224, | |
| "learning_rate": 5.449673790581611e-07, | |
| "loss": 0.5164, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.597444089456869, | |
| "grad_norm": 0.3987196385860443, | |
| "learning_rate": 5.197795201563744e-07, | |
| "loss": 0.5303, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.6070287539936103, | |
| "grad_norm": 0.43403735756874084, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.5029, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.6166134185303513, | |
| "grad_norm": 0.4055974781513214, | |
| "learning_rate": 4.710988998640298e-07, | |
| "loss": 0.5261, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.626198083067093, | |
| "grad_norm": 0.3974541425704956, | |
| "learning_rate": 4.4761226670592074e-07, | |
| "loss": 0.525, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.635782747603834, | |
| "grad_norm": 0.419546514749527, | |
| "learning_rate": 4.2469871766340096e-07, | |
| "loss": 0.5155, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.6453674121405752, | |
| "grad_norm": 0.43064308166503906, | |
| "learning_rate": 4.0236113724274716e-07, | |
| "loss": 0.5245, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.6549520766773163, | |
| "grad_norm": 0.42868760228157043, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.5092, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.6645367412140573, | |
| "grad_norm": 0.4428861737251282, | |
| "learning_rate": 3.5942505740480583e-07, | |
| "loss": 0.5106, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.6741214057507987, | |
| "grad_norm": 0.42494478821754456, | |
| "learning_rate": 3.3883196305992906e-07, | |
| "loss": 0.5048, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.68370607028754, | |
| "grad_norm": 0.4355878531932831, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.5451, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.693290734824281, | |
| "grad_norm": 0.42539212107658386, | |
| "learning_rate": 2.9940862715390483e-07, | |
| "loss": 0.5415, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.702875399361022, | |
| "grad_norm": 0.4196220636367798, | |
| "learning_rate": 2.8058334845816214e-07, | |
| "loss": 0.53, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.7124600638977636, | |
| "grad_norm": 0.404574453830719, | |
| "learning_rate": 2.6235218056235633e-07, | |
| "loss": 0.5235, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.722044728434505, | |
| "grad_norm": 0.4026959538459778, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 0.5296, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.731629392971246, | |
| "grad_norm": 0.39630454778671265, | |
| "learning_rate": 2.276812823220964e-07, | |
| "loss": 0.5248, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.741214057507987, | |
| "grad_norm": 0.4264671504497528, | |
| "learning_rate": 2.1124591657534776e-07, | |
| "loss": 0.5079, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.7507987220447285, | |
| "grad_norm": 0.4322397708892822, | |
| "learning_rate": 1.9541339027450256e-07, | |
| "loss": 0.5086, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.7603833865814695, | |
| "grad_norm": 0.44492900371551514, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.527, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.769968051118211, | |
| "grad_norm": 0.4282550811767578, | |
| "learning_rate": 1.6556475227496816e-07, | |
| "loss": 0.5202, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.779552715654952, | |
| "grad_norm": 0.4308444857597351, | |
| "learning_rate": 1.5155239811656562e-07, | |
| "loss": 0.5402, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.7891373801916934, | |
| "grad_norm": 0.40583133697509766, | |
| "learning_rate": 1.3815039801161723e-07, | |
| "loss": 0.5311, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.7987220447284344, | |
| "grad_norm": 0.39951369166374207, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.5438, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.8083067092651754, | |
| "grad_norm": 0.3861788511276245, | |
| "learning_rate": 1.1318413143740436e-07, | |
| "loss": 0.5327, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.817891373801917, | |
| "grad_norm": 0.4173254668712616, | |
| "learning_rate": 1.0162300788382263e-07, | |
| "loss": 0.5078, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.8274760383386583, | |
| "grad_norm": 0.40433382987976074, | |
| "learning_rate": 9.0678523819408e-08, | |
| "loss": 0.5227, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.8370607028753994, | |
| "grad_norm": 0.410818874835968, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.5124, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.8466453674121404, | |
| "grad_norm": 0.38880231976509094, | |
| "learning_rate": 7.064490740882057e-08, | |
| "loss": 0.5514, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.856230031948882, | |
| "grad_norm": 0.4121662974357605, | |
| "learning_rate": 6.15582970243117e-08, | |
| "loss": 0.5235, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.8658146964856233, | |
| "grad_norm": 0.4136061668395996, | |
| "learning_rate": 5.3093369734816824e-08, | |
| "loss": 0.5121, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.8753993610223643, | |
| "grad_norm": 0.3969224989414215, | |
| "learning_rate": 4.52511911603265e-08, | |
| "loss": 0.5276, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.8849840255591053, | |
| "grad_norm": 0.3935069143772125, | |
| "learning_rate": 3.8032748525179684e-08, | |
| "loss": 0.5266, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.8945686900958467, | |
| "grad_norm": 0.4016058146953583, | |
| "learning_rate": 3.143895053378698e-08, | |
| "loss": 0.5162, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.9041533546325877, | |
| "grad_norm": 0.39917394518852234, | |
| "learning_rate": 2.547062725623828e-08, | |
| "loss": 0.5252, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.913738019169329, | |
| "grad_norm": 0.44093358516693115, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.5069, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.92332268370607, | |
| "grad_norm": 0.39468714594841003, | |
| "learning_rate": 1.541333133436018e-08, | |
| "loss": 0.5177, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.9329073482428116, | |
| "grad_norm": 0.41948553919792175, | |
| "learning_rate": 1.132562476771959e-08, | |
| "loss": 0.5119, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.9424920127795526, | |
| "grad_norm": 0.4047912359237671, | |
| "learning_rate": 7.865924910916977e-09, | |
| "loss": 0.5153, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.952076677316294, | |
| "grad_norm": 0.4035501182079315, | |
| "learning_rate": 5.034667293427053e-09, | |
| "loss": 0.5322, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.961661341853035, | |
| "grad_norm": 0.415784627199173, | |
| "learning_rate": 2.8322083323334417e-09, | |
| "loss": 0.5244, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.9712460063897765, | |
| "grad_norm": 0.40659230947494507, | |
| "learning_rate": 1.2588252874673469e-09, | |
| "loss": 0.508, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.9808306709265175, | |
| "grad_norm": 0.3864118754863739, | |
| "learning_rate": 3.147162264971471e-10, | |
| "loss": 0.5463, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.9904153354632586, | |
| "grad_norm": 0.40317609906196594, | |
| "learning_rate": 0.0, | |
| "loss": 0.5152, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.9904153354632586, | |
| "step": 312, | |
| "total_flos": 3.9082027465048064e+17, | |
| "train_loss": 0.0, | |
| "train_runtime": 6.1081, | |
| "train_samples_per_second": 4911.493, | |
| "train_steps_per_second": 51.08 | |
| } | |
| ], | |
| "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": 3.9082027465048064e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |