| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 1863, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0005367686527106817, | |
| "grad_norm": 23.228396541855062, | |
| "learning_rate": 5.3475935828877005e-08, | |
| "loss": 1.2954, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0026838432635534087, | |
| "grad_norm": 23.37700328838729, | |
| "learning_rate": 2.6737967914438503e-07, | |
| "loss": 1.3268, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.005367686527106817, | |
| "grad_norm": 15.736329824099501, | |
| "learning_rate": 5.347593582887701e-07, | |
| "loss": 1.277, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.008051529790660225, | |
| "grad_norm": 11.846537169848924, | |
| "learning_rate": 8.021390374331551e-07, | |
| "loss": 1.1184, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.010735373054213635, | |
| "grad_norm": 7.237562962480951, | |
| "learning_rate": 1.0695187165775401e-06, | |
| "loss": 1.0127, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.013419216317767043, | |
| "grad_norm": 3.6451360470757406, | |
| "learning_rate": 1.3368983957219254e-06, | |
| "loss": 0.9198, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.01610305958132045, | |
| "grad_norm": 3.5966256067382822, | |
| "learning_rate": 1.6042780748663103e-06, | |
| "loss": 0.8864, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.01878690284487386, | |
| "grad_norm": 3.1185896525313024, | |
| "learning_rate": 1.8716577540106954e-06, | |
| "loss": 0.8521, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.02147074610842727, | |
| "grad_norm": 2.949524183839941, | |
| "learning_rate": 2.1390374331550802e-06, | |
| "loss": 0.8389, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.024154589371980676, | |
| "grad_norm": 2.9787391965084202, | |
| "learning_rate": 2.4064171122994653e-06, | |
| "loss": 0.8274, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.026838432635534086, | |
| "grad_norm": 2.9044939993745578, | |
| "learning_rate": 2.673796791443851e-06, | |
| "loss": 0.8222, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.029522275899087493, | |
| "grad_norm": 2.875106944469434, | |
| "learning_rate": 2.9411764705882355e-06, | |
| "loss": 0.8087, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.0322061191626409, | |
| "grad_norm": 2.885715478237671, | |
| "learning_rate": 3.2085561497326205e-06, | |
| "loss": 0.788, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.03488996242619431, | |
| "grad_norm": 3.1329646507433826, | |
| "learning_rate": 3.4759358288770056e-06, | |
| "loss": 0.7926, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.03757380568974772, | |
| "grad_norm": 3.0523202009751995, | |
| "learning_rate": 3.7433155080213907e-06, | |
| "loss": 0.786, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.040257648953301126, | |
| "grad_norm": 3.0966691004304154, | |
| "learning_rate": 4.010695187165775e-06, | |
| "loss": 0.781, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.04294149221685454, | |
| "grad_norm": 3.0343017912370884, | |
| "learning_rate": 4.2780748663101604e-06, | |
| "loss": 0.7779, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.045625335480407946, | |
| "grad_norm": 3.0955956596018606, | |
| "learning_rate": 4.5454545454545455e-06, | |
| "loss": 0.7488, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.04830917874396135, | |
| "grad_norm": 3.176458200688146, | |
| "learning_rate": 4.812834224598931e-06, | |
| "loss": 0.763, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.05099302200751476, | |
| "grad_norm": 3.347556753069798, | |
| "learning_rate": 5.0802139037433165e-06, | |
| "loss": 0.736, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.05367686527106817, | |
| "grad_norm": 3.3039529608148848, | |
| "learning_rate": 5.347593582887702e-06, | |
| "loss": 0.733, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.05636070853462158, | |
| "grad_norm": 3.0354665469728825, | |
| "learning_rate": 5.614973262032086e-06, | |
| "loss": 0.7225, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.059044551798174985, | |
| "grad_norm": 3.029511417947508, | |
| "learning_rate": 5.882352941176471e-06, | |
| "loss": 0.7317, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.06172839506172839, | |
| "grad_norm": 2.94371742343579, | |
| "learning_rate": 6.149732620320856e-06, | |
| "loss": 0.73, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.0644122383252818, | |
| "grad_norm": 2.996213395797969, | |
| "learning_rate": 6.417112299465241e-06, | |
| "loss": 0.7116, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.06709608158883522, | |
| "grad_norm": 3.043128606666794, | |
| "learning_rate": 6.684491978609626e-06, | |
| "loss": 0.7102, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.06977992485238863, | |
| "grad_norm": 3.0730454811236614, | |
| "learning_rate": 6.951871657754011e-06, | |
| "loss": 0.7115, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.07246376811594203, | |
| "grad_norm": 2.8286084082595786, | |
| "learning_rate": 7.219251336898396e-06, | |
| "loss": 0.7079, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.07514761137949544, | |
| "grad_norm": 3.044935386927364, | |
| "learning_rate": 7.486631016042781e-06, | |
| "loss": 0.7076, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.07783145464304884, | |
| "grad_norm": 3.121121157411506, | |
| "learning_rate": 7.754010695187166e-06, | |
| "loss": 0.7097, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.08051529790660225, | |
| "grad_norm": 2.9396156789665655, | |
| "learning_rate": 8.02139037433155e-06, | |
| "loss": 0.7022, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.08319914117015566, | |
| "grad_norm": 3.1984734757850526, | |
| "learning_rate": 8.288770053475937e-06, | |
| "loss": 0.6979, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.08588298443370908, | |
| "grad_norm": 3.1699915568172403, | |
| "learning_rate": 8.556149732620321e-06, | |
| "loss": 0.7078, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.08856682769726248, | |
| "grad_norm": 2.7691757862228976, | |
| "learning_rate": 8.823529411764707e-06, | |
| "loss": 0.7098, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.09125067096081589, | |
| "grad_norm": 2.997635358915749, | |
| "learning_rate": 9.090909090909091e-06, | |
| "loss": 0.7098, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.0939345142243693, | |
| "grad_norm": 2.9286411286056784, | |
| "learning_rate": 9.358288770053477e-06, | |
| "loss": 0.6985, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.0966183574879227, | |
| "grad_norm": 3.050959040539156, | |
| "learning_rate": 9.625668449197861e-06, | |
| "loss": 0.7022, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.09930220075147611, | |
| "grad_norm": 2.8934512826852568, | |
| "learning_rate": 9.893048128342247e-06, | |
| "loss": 0.6974, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.10198604401502952, | |
| "grad_norm": 2.7718656295813644, | |
| "learning_rate": 9.999920944317846e-06, | |
| "loss": 0.6986, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.10466988727858294, | |
| "grad_norm": 3.3049667103301985, | |
| "learning_rate": 9.99943783531341e-06, | |
| "loss": 0.706, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.10735373054213634, | |
| "grad_norm": 3.527219646711727, | |
| "learning_rate": 9.998515579512446e-06, | |
| "loss": 0.6929, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.11003757380568975, | |
| "grad_norm": 2.798052102567554, | |
| "learning_rate": 9.997154257925199e-06, | |
| "loss": 0.7128, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.11272141706924316, | |
| "grad_norm": 2.7500531455832173, | |
| "learning_rate": 9.995353990129115e-06, | |
| "loss": 0.6862, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.11540526033279656, | |
| "grad_norm": 2.795161162916103, | |
| "learning_rate": 9.99311493425834e-06, | |
| "loss": 0.6903, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.11808910359634997, | |
| "grad_norm": 2.697385063829242, | |
| "learning_rate": 9.99043728698983e-06, | |
| "loss": 0.6805, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.12077294685990338, | |
| "grad_norm": 3.178036882172029, | |
| "learning_rate": 9.987321283526072e-06, | |
| "loss": 0.6743, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.12345679012345678, | |
| "grad_norm": 2.6934310668132118, | |
| "learning_rate": 9.983767197574432e-06, | |
| "loss": 0.681, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.1261406333870102, | |
| "grad_norm": 7.721696196519568, | |
| "learning_rate": 9.979775341323097e-06, | |
| "loss": 0.6805, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.1288244766505636, | |
| "grad_norm": 2.4485392496997793, | |
| "learning_rate": 9.975346065413673e-06, | |
| "loss": 0.6849, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.13150831991411702, | |
| "grad_norm": 2.5711938120298607, | |
| "learning_rate": 9.970479758910365e-06, | |
| "loss": 0.6929, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.13419216317767044, | |
| "grad_norm": 2.722599378322932, | |
| "learning_rate": 9.965176849265814e-06, | |
| "loss": 0.6748, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.13687600644122383, | |
| "grad_norm": 2.5051734527769653, | |
| "learning_rate": 9.959437802283552e-06, | |
| "loss": 0.6725, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.13955984970477725, | |
| "grad_norm": 2.511468365101751, | |
| "learning_rate": 9.953263122077077e-06, | |
| "loss": 0.6697, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.14224369296833064, | |
| "grad_norm": 2.578814660679636, | |
| "learning_rate": 9.946653351025575e-06, | |
| "loss": 0.6707, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.14492753623188406, | |
| "grad_norm": 2.620578729782855, | |
| "learning_rate": 9.939609069726279e-06, | |
| "loss": 0.6715, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.14761137949543746, | |
| "grad_norm": 2.9578009527592397, | |
| "learning_rate": 9.932130896943477e-06, | |
| "loss": 0.6786, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.15029522275899088, | |
| "grad_norm": 2.5757704648830266, | |
| "learning_rate": 9.924219489554145e-06, | |
| "loss": 0.6793, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.1529790660225443, | |
| "grad_norm": 2.5434720920972946, | |
| "learning_rate": 9.915875542490257e-06, | |
| "loss": 0.6592, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.1556629092860977, | |
| "grad_norm": 2.681548893758978, | |
| "learning_rate": 9.907099788677745e-06, | |
| "loss": 0.6604, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.1583467525496511, | |
| "grad_norm": 2.7335381558913308, | |
| "learning_rate": 9.897892998972113e-06, | |
| "loss": 0.6803, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.1610305958132045, | |
| "grad_norm": 2.6325812287223673, | |
| "learning_rate": 9.888255982090728e-06, | |
| "loss": 0.6664, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.16371443907675792, | |
| "grad_norm": 2.511790182600164, | |
| "learning_rate": 9.878189584541783e-06, | |
| "loss": 0.6472, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.16639828234031132, | |
| "grad_norm": 2.6462659551093726, | |
| "learning_rate": 9.867694690549943e-06, | |
| "loss": 0.6625, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.16908212560386474, | |
| "grad_norm": 2.640400429583718, | |
| "learning_rate": 9.85677222197867e-06, | |
| "loss": 0.6517, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.17176596886741816, | |
| "grad_norm": 3.2099275300784855, | |
| "learning_rate": 9.845423138249254e-06, | |
| "loss": 0.6462, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.17444981213097155, | |
| "grad_norm": 2.460040741838709, | |
| "learning_rate": 9.833648436256525e-06, | |
| "loss": 0.6471, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.17713365539452497, | |
| "grad_norm": 2.5168759595887846, | |
| "learning_rate": 9.821449150281308e-06, | |
| "loss": 0.6495, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.17981749865807836, | |
| "grad_norm": 2.3721443436338854, | |
| "learning_rate": 9.808826351899551e-06, | |
| "loss": 0.6697, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.18250134192163178, | |
| "grad_norm": 2.967361065842744, | |
| "learning_rate": 9.795781149888216e-06, | |
| "loss": 0.6677, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.18518518518518517, | |
| "grad_norm": 2.5524417086527356, | |
| "learning_rate": 9.782314690127867e-06, | |
| "loss": 0.6596, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.1878690284487386, | |
| "grad_norm": 2.396536419446281, | |
| "learning_rate": 9.768428155502038e-06, | |
| "loss": 0.6639, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.19055287171229202, | |
| "grad_norm": 2.4081452895879303, | |
| "learning_rate": 9.754122765793306e-06, | |
| "loss": 0.6491, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.1932367149758454, | |
| "grad_norm": 2.560002783053779, | |
| "learning_rate": 9.739399777576169e-06, | |
| "loss": 0.6578, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.19592055823939883, | |
| "grad_norm": 2.4228815507523596, | |
| "learning_rate": 9.72426048410665e-06, | |
| "loss": 0.6321, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.19860440150295222, | |
| "grad_norm": 2.4580386617883763, | |
| "learning_rate": 9.70870621520871e-06, | |
| "loss": 0.6369, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.20128824476650564, | |
| "grad_norm": 2.3874562229297265, | |
| "learning_rate": 9.692738337157441e-06, | |
| "loss": 0.6408, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.20397208803005903, | |
| "grad_norm": 2.5671634163145955, | |
| "learning_rate": 9.676358252559034e-06, | |
| "loss": 0.6449, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.20665593129361245, | |
| "grad_norm": 2.4968314852065587, | |
| "learning_rate": 9.6595674002276e-06, | |
| "loss": 0.6489, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.20933977455716588, | |
| "grad_norm": 2.5121647933384437, | |
| "learning_rate": 9.642367255058767e-06, | |
| "loss": 0.6312, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.21202361782071927, | |
| "grad_norm": 2.369426926693552, | |
| "learning_rate": 9.624759327900131e-06, | |
| "loss": 0.6507, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.2147074610842727, | |
| "grad_norm": 2.4524363201674335, | |
| "learning_rate": 9.606745165418554e-06, | |
| "loss": 0.6311, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.21739130434782608, | |
| "grad_norm": 2.2583158583019043, | |
| "learning_rate": 9.58832634996429e-06, | |
| "loss": 0.6319, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.2200751476113795, | |
| "grad_norm": 2.318154834938378, | |
| "learning_rate": 9.569504499432005e-06, | |
| "loss": 0.6316, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.2227589908749329, | |
| "grad_norm": 2.457228379638502, | |
| "learning_rate": 9.550281267118659e-06, | |
| "loss": 0.6245, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.22544283413848631, | |
| "grad_norm": 2.530926328235379, | |
| "learning_rate": 9.530658341578276e-06, | |
| "loss": 0.6155, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.2281266774020397, | |
| "grad_norm": 2.711138329301352, | |
| "learning_rate": 9.510637446473633e-06, | |
| "loss": 0.6111, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.23081052066559313, | |
| "grad_norm": 2.581880858350562, | |
| "learning_rate": 9.490220340424844e-06, | |
| "loss": 0.6111, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.23349436392914655, | |
| "grad_norm": 2.436448968761118, | |
| "learning_rate": 9.469408816854898e-06, | |
| "loss": 0.6042, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.23617820719269994, | |
| "grad_norm": 2.6980072007155407, | |
| "learning_rate": 9.448204703832102e-06, | |
| "loss": 0.631, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.23886205045625336, | |
| "grad_norm": 2.6629238830814463, | |
| "learning_rate": 9.426609863909537e-06, | |
| "loss": 0.6132, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.24154589371980675, | |
| "grad_norm": 2.363424798307805, | |
| "learning_rate": 9.404626193961427e-06, | |
| "loss": 0.6238, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.24422973698336017, | |
| "grad_norm": 2.6899604011857483, | |
| "learning_rate": 9.382255625016527e-06, | |
| "loss": 0.6192, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.24691358024691357, | |
| "grad_norm": 2.773443910599498, | |
| "learning_rate": 9.359500122088511e-06, | |
| "loss": 0.6123, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.249597423510467, | |
| "grad_norm": 2.4142194950402764, | |
| "learning_rate": 9.336361684003353e-06, | |
| "loss": 0.6191, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.2522812667740204, | |
| "grad_norm": 2.3823560617454858, | |
| "learning_rate": 9.312842343223764e-06, | |
| "loss": 0.5902, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.2549651100375738, | |
| "grad_norm": 2.8064118648442236, | |
| "learning_rate": 9.288944165670651e-06, | |
| "loss": 0.6067, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.2576489533011272, | |
| "grad_norm": 2.453743523372246, | |
| "learning_rate": 9.264669250541658e-06, | |
| "loss": 0.6113, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.2603327965646806, | |
| "grad_norm": 2.3494257863438928, | |
| "learning_rate": 9.240019730126764e-06, | |
| "loss": 0.5979, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.26301663982823403, | |
| "grad_norm": 2.4239444646180455, | |
| "learning_rate": 9.214997769620998e-06, | |
| "loss": 0.5971, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.26570048309178745, | |
| "grad_norm": 2.421915954160417, | |
| "learning_rate": 9.189605566934235e-06, | |
| "loss": 0.5839, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.2683843263553409, | |
| "grad_norm": 2.7119720971165724, | |
| "learning_rate": 9.163845352498141e-06, | |
| "loss": 0.592, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.27106816961889424, | |
| "grad_norm": 2.6152819507580225, | |
| "learning_rate": 9.137719389070259e-06, | |
| "loss": 0.5971, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.27375201288244766, | |
| "grad_norm": 2.511252395129398, | |
| "learning_rate": 9.111229971535231e-06, | |
| "loss": 0.5948, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.2764358561460011, | |
| "grad_norm": 2.352722280092554, | |
| "learning_rate": 9.084379426703245e-06, | |
| "loss": 0.5766, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.2791196994095545, | |
| "grad_norm": 2.341942823394511, | |
| "learning_rate": 9.057170113105622e-06, | |
| "loss": 0.6, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.28180354267310787, | |
| "grad_norm": 2.429806619395087, | |
| "learning_rate": 9.029604420787666e-06, | |
| "loss": 0.5868, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.2844873859366613, | |
| "grad_norm": 2.5206815749367735, | |
| "learning_rate": 9.001684771098709e-06, | |
| "loss": 0.5887, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.2871712292002147, | |
| "grad_norm": 2.3385702289863355, | |
| "learning_rate": 8.973413616479429e-06, | |
| "loss": 0.5764, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.2898550724637681, | |
| "grad_norm": 2.4081982129228887, | |
| "learning_rate": 8.944793440246435e-06, | |
| "loss": 0.5734, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.29253891572732155, | |
| "grad_norm": 2.364008343711155, | |
| "learning_rate": 8.915826756374118e-06, | |
| "loss": 0.5888, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.2952227589908749, | |
| "grad_norm": 2.514918994939067, | |
| "learning_rate": 8.88651610927384e-06, | |
| "loss": 0.5728, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.29790660225442833, | |
| "grad_norm": 4.806599422895776, | |
| "learning_rate": 8.856864073570429e-06, | |
| "loss": 0.5942, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.30059044551798175, | |
| "grad_norm": 2.4038630698543044, | |
| "learning_rate": 8.82687325387603e-06, | |
| "loss": 0.5806, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.3032742887815352, | |
| "grad_norm": 2.281605838359386, | |
| "learning_rate": 8.796546284561307e-06, | |
| "loss": 0.588, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.3059581320450886, | |
| "grad_norm": 2.4124984398821856, | |
| "learning_rate": 8.765885829524059e-06, | |
| "loss": 0.5736, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.30864197530864196, | |
| "grad_norm": 2.400485514024389, | |
| "learning_rate": 8.734894581955208e-06, | |
| "loss": 0.5749, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.3113258185721954, | |
| "grad_norm": 2.3220156657244857, | |
| "learning_rate": 8.703575264102245e-06, | |
| "loss": 0.5643, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.3140096618357488, | |
| "grad_norm": 2.5401487089259787, | |
| "learning_rate": 8.6719306270301e-06, | |
| "loss": 0.5731, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.3166935050993022, | |
| "grad_norm": 2.3210117695521157, | |
| "learning_rate": 8.639963450379494e-06, | |
| "loss": 0.5686, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.3193773483628556, | |
| "grad_norm": 2.6285989359942232, | |
| "learning_rate": 8.607676542122782e-06, | |
| "loss": 0.5594, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.322061191626409, | |
| "grad_norm": 2.767952346585199, | |
| "learning_rate": 8.575072738317297e-06, | |
| "loss": 0.5914, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.3247450348899624, | |
| "grad_norm": 2.584120567147612, | |
| "learning_rate": 8.542154902856232e-06, | |
| "loss": 0.5733, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.32742887815351585, | |
| "grad_norm": 3.141778459777587, | |
| "learning_rate": 8.508925927217083e-06, | |
| "loss": 0.5575, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.33011272141706927, | |
| "grad_norm": 2.3022892848405783, | |
| "learning_rate": 8.475388730207662e-06, | |
| "loss": 0.5591, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.33279656468062263, | |
| "grad_norm": 2.402430056616631, | |
| "learning_rate": 8.441546257709708e-06, | |
| "loss": 0.5732, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.33548040794417605, | |
| "grad_norm": 2.4310246577287713, | |
| "learning_rate": 8.407401482420129e-06, | |
| "loss": 0.545, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.33816425120772947, | |
| "grad_norm": 2.460078290200426, | |
| "learning_rate": 8.372957403589873e-06, | |
| "loss": 0.5559, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.3408480944712829, | |
| "grad_norm": 2.355259854780783, | |
| "learning_rate": 8.33821704676049e-06, | |
| "loss": 0.5597, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.3435319377348363, | |
| "grad_norm": 2.4594517447294053, | |
| "learning_rate": 8.303183463498357e-06, | |
| "loss": 0.5597, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.3462157809983897, | |
| "grad_norm": 2.4249710336123886, | |
| "learning_rate": 8.26785973112664e-06, | |
| "loss": 0.551, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.3488996242619431, | |
| "grad_norm": 2.7442586205687993, | |
| "learning_rate": 8.232248952454978e-06, | |
| "loss": 0.5489, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.3515834675254965, | |
| "grad_norm": 2.5602834292032086, | |
| "learning_rate": 8.196354255506937e-06, | |
| "loss": 0.5505, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.35426731078904994, | |
| "grad_norm": 2.5066605340197814, | |
| "learning_rate": 8.160178793245254e-06, | |
| "loss": 0.5384, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.3569511540526033, | |
| "grad_norm": 2.3497888643625884, | |
| "learning_rate": 8.12372574329487e-06, | |
| "loss": 0.5611, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.3596349973161567, | |
| "grad_norm": 2.410122692303016, | |
| "learning_rate": 8.086998307663815e-06, | |
| "loss": 0.5516, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.36231884057971014, | |
| "grad_norm": 2.259126530084417, | |
| "learning_rate": 8.049999712461956e-06, | |
| "loss": 0.5435, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.36500268384326356, | |
| "grad_norm": 2.50791143732394, | |
| "learning_rate": 8.012733207617602e-06, | |
| "loss": 0.5509, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.367686527106817, | |
| "grad_norm": 2.365716849649686, | |
| "learning_rate": 7.975202066592038e-06, | |
| "loss": 0.5553, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.37037037037037035, | |
| "grad_norm": 2.3759918711677694, | |
| "learning_rate": 7.937409586091996e-06, | |
| "loss": 0.5324, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.37305421363392377, | |
| "grad_norm": 2.472274071710255, | |
| "learning_rate": 7.899359085780062e-06, | |
| "loss": 0.5442, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.3757380568974772, | |
| "grad_norm": 2.41926222864344, | |
| "learning_rate": 7.86105390798309e-06, | |
| "loss": 0.53, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.3784219001610306, | |
| "grad_norm": 2.446732393857762, | |
| "learning_rate": 7.822497417398607e-06, | |
| "loss": 0.5428, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.38110574342458403, | |
| "grad_norm": 2.5447746479127034, | |
| "learning_rate": 7.783693000799263e-06, | |
| "loss": 0.5419, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.3837895866881374, | |
| "grad_norm": 2.6086841856188894, | |
| "learning_rate": 7.744644066735335e-06, | |
| "loss": 0.5334, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.3864734299516908, | |
| "grad_norm": 2.5629007671927577, | |
| "learning_rate": 7.705354045235335e-06, | |
| "loss": 0.5353, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.38915727321524424, | |
| "grad_norm": 2.280032376143978, | |
| "learning_rate": 7.665826387504708e-06, | |
| "loss": 0.5333, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.39184111647879766, | |
| "grad_norm": 2.3109664603252296, | |
| "learning_rate": 7.626064565622684e-06, | |
| "loss": 0.5151, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.394524959742351, | |
| "grad_norm": 2.6257222367728477, | |
| "learning_rate": 7.586072072237291e-06, | |
| "loss": 0.5346, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.39720880300590444, | |
| "grad_norm": 2.4165458715622252, | |
| "learning_rate": 7.545852420258566e-06, | |
| "loss": 0.5285, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.39989264626945786, | |
| "grad_norm": 2.2702828710603358, | |
| "learning_rate": 7.505409142549987e-06, | |
| "loss": 0.5163, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.4025764895330113, | |
| "grad_norm": 2.399275682318177, | |
| "learning_rate": 7.464745791618138e-06, | |
| "loss": 0.5219, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.4052603327965647, | |
| "grad_norm": 2.229762579994072, | |
| "learning_rate": 7.423865939300674e-06, | |
| "loss": 0.5209, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.40794417606011807, | |
| "grad_norm": 2.534982859498352, | |
| "learning_rate": 7.382773176452561e-06, | |
| "loss": 0.5118, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.4106280193236715, | |
| "grad_norm": 2.496270283095597, | |
| "learning_rate": 7.341471112630667e-06, | |
| "loss": 0.4952, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.4133118625872249, | |
| "grad_norm": 2.256598576541155, | |
| "learning_rate": 7.2999633757766956e-06, | |
| "loss": 0.5139, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.41599570585077833, | |
| "grad_norm": 2.4653796883168217, | |
| "learning_rate": 7.258253611898509e-06, | |
| "loss": 0.5095, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.41867954911433175, | |
| "grad_norm": 2.3552712367439095, | |
| "learning_rate": 7.216345484749876e-06, | |
| "loss": 0.5011, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.4213633923778851, | |
| "grad_norm": 2.329715385386875, | |
| "learning_rate": 7.174242675508636e-06, | |
| "loss": 0.5101, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.42404723564143854, | |
| "grad_norm": 2.4483716935458486, | |
| "learning_rate": 7.131948882453361e-06, | |
| "loss": 0.5161, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.42673107890499196, | |
| "grad_norm": 2.293644946763117, | |
| "learning_rate": 7.089467820638491e-06, | |
| "loss": 0.5039, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.4294149221685454, | |
| "grad_norm": 2.6328009026942802, | |
| "learning_rate": 7.0468032215680116e-06, | |
| "loss": 0.5139, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.43209876543209874, | |
| "grad_norm": 2.4390910161191397, | |
| "learning_rate": 7.003958832867681e-06, | |
| "loss": 0.5074, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.43478260869565216, | |
| "grad_norm": 2.283895276345824, | |
| "learning_rate": 6.960938417955841e-06, | |
| "loss": 0.51, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.4374664519592056, | |
| "grad_norm": 2.303157334285009, | |
| "learning_rate": 6.917745755712839e-06, | |
| "loss": 0.4997, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.440150295222759, | |
| "grad_norm": 2.4059405395054134, | |
| "learning_rate": 6.874384640149101e-06, | |
| "loss": 0.4958, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.4428341384863124, | |
| "grad_norm": 2.317723718127889, | |
| "learning_rate": 6.830858880071859e-06, | |
| "loss": 0.5088, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.4455179817498658, | |
| "grad_norm": 2.338645412216065, | |
| "learning_rate": 6.7871722987505975e-06, | |
| "loss": 0.4948, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.4482018250134192, | |
| "grad_norm": 2.3798351295711093, | |
| "learning_rate": 6.743328733581211e-06, | |
| "loss": 0.5131, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.45088566827697263, | |
| "grad_norm": 2.480685418217814, | |
| "learning_rate": 6.6993320357489396e-06, | |
| "loss": 0.5078, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.45356951154052605, | |
| "grad_norm": 2.467672744334298, | |
| "learning_rate": 6.6551860698900804e-06, | |
| "loss": 0.5002, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.4562533548040794, | |
| "grad_norm": 2.2968442187003544, | |
| "learning_rate": 6.610894713752513e-06, | |
| "loss": 0.5011, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.45893719806763283, | |
| "grad_norm": 2.412385740094048, | |
| "learning_rate": 6.566461857855096e-06, | |
| "loss": 0.4781, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.46162104133118625, | |
| "grad_norm": 2.323965692766263, | |
| "learning_rate": 6.52189140514591e-06, | |
| "loss": 0.4881, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.4643048845947397, | |
| "grad_norm": 2.2689915089515598, | |
| "learning_rate": 6.477187270659444e-06, | |
| "loss": 0.4824, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.4669887278582931, | |
| "grad_norm": 2.3382442657756406, | |
| "learning_rate": 6.432353381172686e-06, | |
| "loss": 0.4682, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.46967257112184646, | |
| "grad_norm": 2.4181856369825496, | |
| "learning_rate": 6.387393674860205e-06, | |
| "loss": 0.4801, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.4723564143853999, | |
| "grad_norm": 2.4127430052478194, | |
| "learning_rate": 6.3423121009482226e-06, | |
| "loss": 0.4836, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.4750402576489533, | |
| "grad_norm": 2.3426581165871263, | |
| "learning_rate": 6.297112619367721e-06, | |
| "loss": 0.4784, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.4777241009125067, | |
| "grad_norm": 2.483246104308074, | |
| "learning_rate": 6.2517992004066e-06, | |
| "loss": 0.4817, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.48040794417606014, | |
| "grad_norm": 2.3477050009099782, | |
| "learning_rate": 6.2063758243609275e-06, | |
| "loss": 0.4891, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.4830917874396135, | |
| "grad_norm": 2.260155765873621, | |
| "learning_rate": 6.160846481185326e-06, | |
| "loss": 0.4675, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.4857756307031669, | |
| "grad_norm": 2.387383273521845, | |
| "learning_rate": 6.115215170142481e-06, | |
| "loss": 0.4974, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.48845947396672035, | |
| "grad_norm": 2.438771805892275, | |
| "learning_rate": 6.069485899451863e-06, | |
| "loss": 0.4959, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.49114331723027377, | |
| "grad_norm": 2.411697824531355, | |
| "learning_rate": 6.023662685937643e-06, | |
| "loss": 0.4794, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.49382716049382713, | |
| "grad_norm": 2.3344015335242667, | |
| "learning_rate": 5.9777495546758556e-06, | |
| "loss": 0.4586, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.49651100375738055, | |
| "grad_norm": 2.357939860685128, | |
| "learning_rate": 5.9317505386408436e-06, | |
| "loss": 0.4693, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.499194847020934, | |
| "grad_norm": 2.273317709039976, | |
| "learning_rate": 5.885669678350997e-06, | |
| "loss": 0.4707, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.5018786902844874, | |
| "grad_norm": 2.339575461889168, | |
| "learning_rate": 5.839511021513853e-06, | |
| "loss": 0.47, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.5045625335480408, | |
| "grad_norm": 2.4281112555555775, | |
| "learning_rate": 5.793278622670527e-06, | |
| "loss": 0.4745, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.5072463768115942, | |
| "grad_norm": 2.3474904478592116, | |
| "learning_rate": 5.746976542839583e-06, | |
| "loss": 0.4636, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.5099302200751477, | |
| "grad_norm": 2.3520486189119643, | |
| "learning_rate": 5.700608849160307e-06, | |
| "loss": 0.4743, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.512614063338701, | |
| "grad_norm": 2.395554773244391, | |
| "learning_rate": 5.654179614535457e-06, | |
| "loss": 0.4551, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.5152979066022544, | |
| "grad_norm": 2.275110339749561, | |
| "learning_rate": 5.607692917273499e-06, | |
| "loss": 0.4458, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.5179817498658078, | |
| "grad_norm": 2.421123904254612, | |
| "learning_rate": 5.561152840730371e-06, | |
| "loss": 0.4711, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.5206655931293612, | |
| "grad_norm": 2.2497808537496713, | |
| "learning_rate": 5.514563472950812e-06, | |
| "loss": 0.4696, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.5233494363929146, | |
| "grad_norm": 2.5178182576633903, | |
| "learning_rate": 5.46792890630926e-06, | |
| "loss": 0.4509, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.5260332796564681, | |
| "grad_norm": 2.224586181033126, | |
| "learning_rate": 5.421253237150389e-06, | |
| "loss": 0.4616, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.5287171229200215, | |
| "grad_norm": 2.295379835377096, | |
| "learning_rate": 5.374540565429288e-06, | |
| "loss": 0.4541, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.5314009661835749, | |
| "grad_norm": 2.3765802872433945, | |
| "learning_rate": 5.327794994351321e-06, | |
| "loss": 0.4598, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.5340848094471283, | |
| "grad_norm": 2.187775346368912, | |
| "learning_rate": 5.281020630011703e-06, | |
| "loss": 0.4445, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.5367686527106817, | |
| "grad_norm": 2.1958208915497544, | |
| "learning_rate": 5.234221581034834e-06, | |
| "loss": 0.4561, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.5394524959742351, | |
| "grad_norm": 2.3014363536164715, | |
| "learning_rate": 5.187401958213391e-06, | |
| "loss": 0.4649, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.5421363392377885, | |
| "grad_norm": 2.227842407723195, | |
| "learning_rate": 5.140565874147236e-06, | |
| "loss": 0.4605, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.5448201825013419, | |
| "grad_norm": 2.38442176913314, | |
| "learning_rate": 5.093717442882185e-06, | |
| "loss": 0.4379, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.5475040257648953, | |
| "grad_norm": 2.2271764971217847, | |
| "learning_rate": 5.046860779548613e-06, | |
| "loss": 0.426, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.5501878690284487, | |
| "grad_norm": 2.3744379994527267, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4602, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.5528717122920022, | |
| "grad_norm": 2.6385293936056486, | |
| "learning_rate": 4.953139220451386e-06, | |
| "loss": 0.4455, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 2.2086213112270072, | |
| "learning_rate": 4.906282557117817e-06, | |
| "loss": 0.4393, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.558239398819109, | |
| "grad_norm": 2.3995862592671044, | |
| "learning_rate": 4.859434125852764e-06, | |
| "loss": 0.4392, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.5609232420826624, | |
| "grad_norm": 2.4911947290870335, | |
| "learning_rate": 4.812598041786611e-06, | |
| "loss": 0.4423, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.5636070853462157, | |
| "grad_norm": 2.411600620645279, | |
| "learning_rate": 4.765778418965168e-06, | |
| "loss": 0.4348, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.5662909286097692, | |
| "grad_norm": 2.595675161221087, | |
| "learning_rate": 4.718979369988299e-06, | |
| "loss": 0.4449, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.5689747718733226, | |
| "grad_norm": 2.2951677469132274, | |
| "learning_rate": 4.672205005648681e-06, | |
| "loss": 0.436, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.571658615136876, | |
| "grad_norm": 2.2713867895597515, | |
| "learning_rate": 4.625459434570712e-06, | |
| "loss": 0.4317, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.5743424584004294, | |
| "grad_norm": 2.1919155354800672, | |
| "learning_rate": 4.578746762849612e-06, | |
| "loss": 0.4147, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.5770263016639828, | |
| "grad_norm": 2.191444357915981, | |
| "learning_rate": 4.532071093690741e-06, | |
| "loss": 0.4192, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.5797101449275363, | |
| "grad_norm": 2.285151573730604, | |
| "learning_rate": 4.485436527049189e-06, | |
| "loss": 0.4307, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.5823939881910897, | |
| "grad_norm": 2.171625114460924, | |
| "learning_rate": 4.438847159269631e-06, | |
| "loss": 0.424, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.5850778314546431, | |
| "grad_norm": 2.3044237354743347, | |
| "learning_rate": 4.392307082726503e-06, | |
| "loss": 0.4348, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.5877616747181964, | |
| "grad_norm": 2.2777977939526, | |
| "learning_rate": 4.345820385464543e-06, | |
| "loss": 0.433, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.5904455179817498, | |
| "grad_norm": 2.324478594910565, | |
| "learning_rate": 4.299391150839694e-06, | |
| "loss": 0.4349, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.5931293612453032, | |
| "grad_norm": 2.331352000617292, | |
| "learning_rate": 4.253023457160418e-06, | |
| "loss": 0.4273, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.5958132045088567, | |
| "grad_norm": 2.205631106966491, | |
| "learning_rate": 4.2067213773294745e-06, | |
| "loss": 0.4308, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.5984970477724101, | |
| "grad_norm": 2.2955340806741775, | |
| "learning_rate": 4.16048897848615e-06, | |
| "loss": 0.4163, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.6011808910359635, | |
| "grad_norm": 2.277025683681091, | |
| "learning_rate": 4.114330321649004e-06, | |
| "loss": 0.4178, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.6038647342995169, | |
| "grad_norm": 2.2989398993559784, | |
| "learning_rate": 4.068249461359158e-06, | |
| "loss": 0.4223, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.6065485775630703, | |
| "grad_norm": 2.2434708014272107, | |
| "learning_rate": 4.022250445324145e-06, | |
| "loss": 0.4121, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.6092324208266238, | |
| "grad_norm": 2.353097002584675, | |
| "learning_rate": 3.976337314062358e-06, | |
| "loss": 0.4098, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 0.6119162640901772, | |
| "grad_norm": 2.2441818809347462, | |
| "learning_rate": 3.9305141005481385e-06, | |
| "loss": 0.412, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.6146001073537305, | |
| "grad_norm": 2.2010756089398456, | |
| "learning_rate": 3.8847848298575205e-06, | |
| "loss": 0.4163, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 0.6172839506172839, | |
| "grad_norm": 2.16029855909458, | |
| "learning_rate": 3.839153518814677e-06, | |
| "loss": 0.414, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.6199677938808373, | |
| "grad_norm": 2.322268458535899, | |
| "learning_rate": 3.7936241756390746e-06, | |
| "loss": 0.4202, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.6226516371443908, | |
| "grad_norm": 2.275309278083409, | |
| "learning_rate": 3.748200799593402e-06, | |
| "loss": 0.4155, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.6253354804079442, | |
| "grad_norm": 2.2513350482004535, | |
| "learning_rate": 3.7028873806322797e-06, | |
| "loss": 0.4183, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 0.6280193236714976, | |
| "grad_norm": 2.214892864993843, | |
| "learning_rate": 3.6576878990517783e-06, | |
| "loss": 0.3981, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.630703166935051, | |
| "grad_norm": 2.2693115523759895, | |
| "learning_rate": 3.6126063251397968e-06, | |
| "loss": 0.4223, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.6333870101986044, | |
| "grad_norm": 2.1004728356465643, | |
| "learning_rate": 3.567646618827315e-06, | |
| "loss": 0.3956, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.6360708534621579, | |
| "grad_norm": 2.199898454082599, | |
| "learning_rate": 3.522812729340557e-06, | |
| "loss": 0.4091, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 0.6387546967257112, | |
| "grad_norm": 2.3531288930136847, | |
| "learning_rate": 3.4781085948540907e-06, | |
| "loss": 0.4126, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.6414385399892646, | |
| "grad_norm": 2.045506572217596, | |
| "learning_rate": 3.4335381421449056e-06, | |
| "loss": 0.3932, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 0.644122383252818, | |
| "grad_norm": 2.2963831027859745, | |
| "learning_rate": 3.3891052862474884e-06, | |
| "loss": 0.4047, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.6468062265163714, | |
| "grad_norm": 2.3796072153277135, | |
| "learning_rate": 3.3448139301099212e-06, | |
| "loss": 0.407, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 0.6494900697799249, | |
| "grad_norm": 2.169493728855937, | |
| "learning_rate": 3.3006679642510596e-06, | |
| "loss": 0.4056, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.6521739130434783, | |
| "grad_norm": 2.282360801451553, | |
| "learning_rate": 3.2566712664187907e-06, | |
| "loss": 0.3965, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 0.6548577563070317, | |
| "grad_norm": 2.3027477380206767, | |
| "learning_rate": 3.2128277012494046e-06, | |
| "loss": 0.4021, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.6575415995705851, | |
| "grad_norm": 2.278369058082134, | |
| "learning_rate": 3.169141119928142e-06, | |
| "loss": 0.4171, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.6602254428341385, | |
| "grad_norm": 2.024962717176371, | |
| "learning_rate": 3.125615359850902e-06, | |
| "loss": 0.3994, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.6629092860976918, | |
| "grad_norm": 2.209336240965979, | |
| "learning_rate": 3.082254244287163e-06, | |
| "loss": 0.4039, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 0.6655931293612453, | |
| "grad_norm": 2.3317565078880813, | |
| "learning_rate": 3.0390615820441605e-06, | |
| "loss": 0.3894, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.6682769726247987, | |
| "grad_norm": 2.290882900173151, | |
| "learning_rate": 2.9960411671323215e-06, | |
| "loss": 0.3963, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 0.6709608158883521, | |
| "grad_norm": 2.3138099906015834, | |
| "learning_rate": 2.9531967784319897e-06, | |
| "loss": 0.3953, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.6736446591519055, | |
| "grad_norm": 2.299501555998995, | |
| "learning_rate": 2.9105321793615106e-06, | |
| "loss": 0.3949, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 0.6763285024154589, | |
| "grad_norm": 2.326104820914772, | |
| "learning_rate": 2.8680511175466384e-06, | |
| "loss": 0.3829, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.6790123456790124, | |
| "grad_norm": 2.2287838569578056, | |
| "learning_rate": 2.8257573244913637e-06, | |
| "loss": 0.402, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 0.6816961889425658, | |
| "grad_norm": 2.207455745073444, | |
| "learning_rate": 2.783654515250126e-06, | |
| "loss": 0.4008, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.6843800322061192, | |
| "grad_norm": 2.236532058684441, | |
| "learning_rate": 2.741746388101493e-06, | |
| "loss": 0.3872, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.6870638754696726, | |
| "grad_norm": 2.17083053593897, | |
| "learning_rate": 2.700036624223308e-06, | |
| "loss": 0.3995, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.6897477187332259, | |
| "grad_norm": 2.2987559681055227, | |
| "learning_rate": 2.6585288873693355e-06, | |
| "loss": 0.3801, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 0.6924315619967794, | |
| "grad_norm": 2.1404443288267925, | |
| "learning_rate": 2.6172268235474396e-06, | |
| "loss": 0.4023, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.6951154052603328, | |
| "grad_norm": 2.2552496716061663, | |
| "learning_rate": 2.576134060699328e-06, | |
| "loss": 0.3936, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 0.6977992485238862, | |
| "grad_norm": 2.298007718871017, | |
| "learning_rate": 2.5352542083818636e-06, | |
| "loss": 0.3815, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.7004830917874396, | |
| "grad_norm": 2.093531968001759, | |
| "learning_rate": 2.494590857450016e-06, | |
| "loss": 0.387, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 0.703166935050993, | |
| "grad_norm": 2.3054180776770075, | |
| "learning_rate": 2.454147579741436e-06, | |
| "loss": 0.3779, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.7058507783145465, | |
| "grad_norm": 2.16129419126451, | |
| "learning_rate": 2.4139279277627113e-06, | |
| "loss": 0.3797, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 0.7085346215780999, | |
| "grad_norm": 2.201892507539766, | |
| "learning_rate": 2.373935434377316e-06, | |
| "loss": 0.3754, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.7112184648416533, | |
| "grad_norm": 2.0925387424217305, | |
| "learning_rate": 2.3341736124952924e-06, | |
| "loss": 0.371, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.7139023081052066, | |
| "grad_norm": 2.1345414088965384, | |
| "learning_rate": 2.294645954764666e-06, | |
| "loss": 0.3742, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.71658615136876, | |
| "grad_norm": 2.2967181107869346, | |
| "learning_rate": 2.2553559332646675e-06, | |
| "loss": 0.378, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 0.7192699946323134, | |
| "grad_norm": 2.288948323295151, | |
| "learning_rate": 2.216306999200739e-06, | |
| "loss": 0.3833, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.7219538378958669, | |
| "grad_norm": 2.1147090940725284, | |
| "learning_rate": 2.1775025826013953e-06, | |
| "loss": 0.366, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 0.7246376811594203, | |
| "grad_norm": 2.514135425876648, | |
| "learning_rate": 2.1389460920169105e-06, | |
| "loss": 0.3778, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.7273215244229737, | |
| "grad_norm": 2.1457736404179495, | |
| "learning_rate": 2.100640914219939e-06, | |
| "loss": 0.383, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 0.7300053676865271, | |
| "grad_norm": 2.2410651034726543, | |
| "learning_rate": 2.062590413908007e-06, | |
| "loss": 0.3753, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.7326892109500805, | |
| "grad_norm": 2.164563325660788, | |
| "learning_rate": 2.0247979334079646e-06, | |
| "loss": 0.3658, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 0.735373054213634, | |
| "grad_norm": 2.0595359788084626, | |
| "learning_rate": 1.9872667923823996e-06, | |
| "loss": 0.3689, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.7380568974771873, | |
| "grad_norm": 2.2887637370315024, | |
| "learning_rate": 1.9500002875380458e-06, | |
| "loss": 0.3756, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "grad_norm": 2.183407906601478, | |
| "learning_rate": 1.9130016923361843e-06, | |
| "loss": 0.3713, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.7434245840042941, | |
| "grad_norm": 2.2015294775680596, | |
| "learning_rate": 1.8762742567051318e-06, | |
| "loss": 0.3746, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 0.7461084272678475, | |
| "grad_norm": 2.193848953829863, | |
| "learning_rate": 1.8398212067547455e-06, | |
| "loss": 0.3712, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.748792270531401, | |
| "grad_norm": 2.199182298510069, | |
| "learning_rate": 1.8036457444930643e-06, | |
| "loss": 0.3637, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 0.7514761137949544, | |
| "grad_norm": 2.177933430454942, | |
| "learning_rate": 1.7677510475450244e-06, | |
| "loss": 0.3719, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.7541599570585078, | |
| "grad_norm": 2.0880489816414123, | |
| "learning_rate": 1.7321402688733618e-06, | |
| "loss": 0.3564, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 0.7568438003220612, | |
| "grad_norm": 2.2027201264261245, | |
| "learning_rate": 1.6968165365016443e-06, | |
| "loss": 0.3702, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.7595276435856146, | |
| "grad_norm": 2.0423356983295307, | |
| "learning_rate": 1.661782953239512e-06, | |
| "loss": 0.3752, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 0.7622114868491681, | |
| "grad_norm": 2.26444271903091, | |
| "learning_rate": 1.6270425964101272e-06, | |
| "loss": 0.3654, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.7648953301127214, | |
| "grad_norm": 2.1770667501287773, | |
| "learning_rate": 1.5925985175798741e-06, | |
| "loss": 0.3578, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.7675791733762748, | |
| "grad_norm": 2.120748383346565, | |
| "learning_rate": 1.5584537422902923e-06, | |
| "loss": 0.3589, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.7702630166398282, | |
| "grad_norm": 2.1144578440481308, | |
| "learning_rate": 1.5246112697923389e-06, | |
| "loss": 0.3639, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 0.7729468599033816, | |
| "grad_norm": 2.165933934760589, | |
| "learning_rate": 1.4910740727829176e-06, | |
| "loss": 0.3626, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.775630703166935, | |
| "grad_norm": 2.194982345171772, | |
| "learning_rate": 1.457845097143769e-06, | |
| "loss": 0.3606, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 0.7783145464304885, | |
| "grad_norm": 1.9892428088264433, | |
| "learning_rate": 1.424927261682703e-06, | |
| "loss": 0.3556, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.7809983896940419, | |
| "grad_norm": 2.352522609198032, | |
| "learning_rate": 1.3923234578772177e-06, | |
| "loss": 0.3663, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 0.7836822329575953, | |
| "grad_norm": 2.03329437167459, | |
| "learning_rate": 1.3600365496205059e-06, | |
| "loss": 0.3608, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.7863660762211487, | |
| "grad_norm": 2.079869447413616, | |
| "learning_rate": 1.3280693729699018e-06, | |
| "loss": 0.3598, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 0.789049919484702, | |
| "grad_norm": 2.1852862851731634, | |
| "learning_rate": 1.2964247358977567e-06, | |
| "loss": 0.3547, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.7917337627482555, | |
| "grad_norm": 2.063384544130908, | |
| "learning_rate": 1.265105418044793e-06, | |
| "loss": 0.3616, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.7944176060118089, | |
| "grad_norm": 2.22988975561877, | |
| "learning_rate": 1.2341141704759418e-06, | |
| "loss": 0.3616, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.7971014492753623, | |
| "grad_norm": 2.1349737925819943, | |
| "learning_rate": 1.2034537154386933e-06, | |
| "loss": 0.3529, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 0.7997852925389157, | |
| "grad_norm": 2.245695097713672, | |
| "learning_rate": 1.173126746123972e-06, | |
| "loss": 0.3541, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.8024691358024691, | |
| "grad_norm": 2.095867298577095, | |
| "learning_rate": 1.1431359264295717e-06, | |
| "loss": 0.3575, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 0.8051529790660226, | |
| "grad_norm": 2.0726227769023895, | |
| "learning_rate": 1.113483890726162e-06, | |
| "loss": 0.3573, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.807836822329576, | |
| "grad_norm": 2.13384295639218, | |
| "learning_rate": 1.084173243625884e-06, | |
| "loss": 0.3593, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 0.8105206655931294, | |
| "grad_norm": 1.957713954201582, | |
| "learning_rate": 1.0552065597535671e-06, | |
| "loss": 0.3523, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.8132045088566827, | |
| "grad_norm": 1.9604899726917404, | |
| "learning_rate": 1.0265863835205709e-06, | |
| "loss": 0.3442, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 0.8158883521202361, | |
| "grad_norm": 2.1648545588299695, | |
| "learning_rate": 9.983152289012926e-07, | |
| "loss": 0.3631, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.8185721953837896, | |
| "grad_norm": 2.0062507922876605, | |
| "learning_rate": 9.70395579212336e-07, | |
| "loss": 0.3455, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.821256038647343, | |
| "grad_norm": 2.071569592380336, | |
| "learning_rate": 9.428298868943791e-07, | |
| "loss": 0.3562, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.8239398819108964, | |
| "grad_norm": 2.1129213446901134, | |
| "learning_rate": 9.15620573296756e-07, | |
| "loss": 0.3574, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 0.8266237251744498, | |
| "grad_norm": 2.1077752969008725, | |
| "learning_rate": 8.887700284647699e-07, | |
| "loss": 0.3521, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.8293075684380032, | |
| "grad_norm": 2.0664306554554015, | |
| "learning_rate": 8.622806109297432e-07, | |
| "loss": 0.3444, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 0.8319914117015567, | |
| "grad_norm": 2.033928416097634, | |
| "learning_rate": 8.361546475018589e-07, | |
| "loss": 0.3431, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.8346752549651101, | |
| "grad_norm": 1.9589716090899654, | |
| "learning_rate": 8.103944330657665e-07, | |
| "loss": 0.3407, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 0.8373590982286635, | |
| "grad_norm": 2.1389604867917327, | |
| "learning_rate": 7.850022303790033e-07, | |
| "loss": 0.3451, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.8400429414922168, | |
| "grad_norm": 2.026518781486662, | |
| "learning_rate": 7.599802698732356e-07, | |
| "loss": 0.338, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 0.8427267847557702, | |
| "grad_norm": 1.996188619746954, | |
| "learning_rate": 7.353307494583445e-07, | |
| "loss": 0.3443, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.8454106280193237, | |
| "grad_norm": 2.1584322204679633, | |
| "learning_rate": 7.1105583432935e-07, | |
| "loss": 0.343, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 0.8480944712828771, | |
| "grad_norm": 2.1866152079241266, | |
| "learning_rate": 6.871576567762373e-07, | |
| "loss": 0.3578, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.8507783145464305, | |
| "grad_norm": 2.0789670521402224, | |
| "learning_rate": 6.63638315996647e-07, | |
| "loss": 0.3457, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 0.8534621578099839, | |
| "grad_norm": 2.04659757991257, | |
| "learning_rate": 6.404998779114912e-07, | |
| "loss": 0.3406, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.8561460010735373, | |
| "grad_norm": 2.0463796960288505, | |
| "learning_rate": 6.177443749834743e-07, | |
| "loss": 0.3389, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 0.8588298443370908, | |
| "grad_norm": 2.1348501710422276, | |
| "learning_rate": 5.953738060385761e-07, | |
| "loss": 0.346, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.8615136876006442, | |
| "grad_norm": 1.9890631177246398, | |
| "learning_rate": 5.733901360904648e-07, | |
| "loss": 0.3479, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 0.8641975308641975, | |
| "grad_norm": 2.1542967746984476, | |
| "learning_rate": 5.517952961678997e-07, | |
| "loss": 0.3453, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.8668813741277509, | |
| "grad_norm": 2.013538176889642, | |
| "learning_rate": 5.305911831451044e-07, | |
| "loss": 0.3502, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "grad_norm": 1.987360819033251, | |
| "learning_rate": 5.097796595751564e-07, | |
| "loss": 0.3412, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.8722490606548577, | |
| "grad_norm": 2.1530291270929727, | |
| "learning_rate": 4.89362553526368e-07, | |
| "loss": 0.3503, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 0.8749329039184112, | |
| "grad_norm": 2.068615177013079, | |
| "learning_rate": 4.6934165842172476e-07, | |
| "loss": 0.3345, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.8776167471819646, | |
| "grad_norm": 1.9948414693248158, | |
| "learning_rate": 4.4971873288134237e-07, | |
| "loss": 0.3406, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 0.880300590445518, | |
| "grad_norm": 2.11263119372112, | |
| "learning_rate": 4.304955005679962e-07, | |
| "loss": 0.3518, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.8829844337090714, | |
| "grad_norm": 1.9707021147537835, | |
| "learning_rate": 4.1167365003571047e-07, | |
| "loss": 0.3417, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 0.8856682769726248, | |
| "grad_norm": 2.196120501296894, | |
| "learning_rate": 3.9325483458144684e-07, | |
| "loss": 0.3423, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.8883521202361782, | |
| "grad_norm": 2.159986760501864, | |
| "learning_rate": 3.752406720998691e-07, | |
| "loss": 0.3532, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 0.8910359634997316, | |
| "grad_norm": 2.002653237944595, | |
| "learning_rate": 3.5763274494123424e-07, | |
| "loss": 0.3465, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.893719806763285, | |
| "grad_norm": 2.090836580626659, | |
| "learning_rate": 3.404325997723995e-07, | |
| "loss": 0.3414, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 0.8964036500268384, | |
| "grad_norm": 1.9607981418295786, | |
| "learning_rate": 3.2364174744096587e-07, | |
| "loss": 0.3415, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.8990874932903918, | |
| "grad_norm": 1.896445035677668, | |
| "learning_rate": 3.072616628425601e-07, | |
| "loss": 0.332, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 0.9017713365539453, | |
| "grad_norm": 2.0613932973395004, | |
| "learning_rate": 2.9129378479129e-07, | |
| "loss": 0.3474, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.9044551798174987, | |
| "grad_norm": 2.257386920061092, | |
| "learning_rate": 2.7573951589335177e-07, | |
| "loss": 0.3488, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 0.9071390230810521, | |
| "grad_norm": 1.950175352619762, | |
| "learning_rate": 2.60600222423833e-07, | |
| "loss": 0.3433, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.9098228663446055, | |
| "grad_norm": 2.091564621614645, | |
| "learning_rate": 2.45877234206694e-07, | |
| "loss": 0.3452, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 0.9125067096081588, | |
| "grad_norm": 1.9669554627425123, | |
| "learning_rate": 2.3157184449796365e-07, | |
| "loss": 0.3392, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.9151905528717122, | |
| "grad_norm": 2.023359207947306, | |
| "learning_rate": 2.1768530987213267e-07, | |
| "loss": 0.334, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 0.9178743961352657, | |
| "grad_norm": 2.1257479374551305, | |
| "learning_rate": 2.0421885011178532e-07, | |
| "loss": 0.333, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.9205582393988191, | |
| "grad_norm": 2.0153576375262214, | |
| "learning_rate": 1.911736481004489e-07, | |
| "loss": 0.3358, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 0.9232420826623725, | |
| "grad_norm": 2.043073503575971, | |
| "learning_rate": 1.7855084971869385e-07, | |
| "loss": 0.3382, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.9259259259259259, | |
| "grad_norm": 2.0328816515974344, | |
| "learning_rate": 1.663515637434765e-07, | |
| "loss": 0.3404, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 0.9286097691894794, | |
| "grad_norm": 1.976185273260926, | |
| "learning_rate": 1.5457686175074816e-07, | |
| "loss": 0.3313, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.9312936124530328, | |
| "grad_norm": 1.7950586305990273, | |
| "learning_rate": 1.432277780213298e-07, | |
| "loss": 0.3393, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 0.9339774557165862, | |
| "grad_norm": 2.198262671444542, | |
| "learning_rate": 1.3230530945005704e-07, | |
| "loss": 0.3365, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.9366612989801396, | |
| "grad_norm": 1.9309012624586674, | |
| "learning_rate": 1.2181041545821727e-07, | |
| "loss": 0.3427, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 0.9393451422436929, | |
| "grad_norm": 1.9610702241447848, | |
| "learning_rate": 1.1174401790927325e-07, | |
| "loss": 0.3367, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.9420289855072463, | |
| "grad_norm": 2.1120678832262305, | |
| "learning_rate": 1.0210700102788796e-07, | |
| "loss": 0.3331, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 0.9447128287707998, | |
| "grad_norm": 1.989237799986007, | |
| "learning_rate": 9.290021132225546e-08, | |
| "loss": 0.3373, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.9473966720343532, | |
| "grad_norm": 1.9299371138740664, | |
| "learning_rate": 8.412445750974385e-08, | |
| "loss": 0.3329, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 0.9500805152979066, | |
| "grad_norm": 2.03429917964713, | |
| "learning_rate": 7.578051044585644e-08, | |
| "loss": 0.3351, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.95276435856146, | |
| "grad_norm": 2.1471433439202703, | |
| "learning_rate": 6.786910305652373e-08, | |
| "loss": 0.3388, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 0.9554482018250134, | |
| "grad_norm": 1.9918329505861243, | |
| "learning_rate": 6.0390930273721e-08, | |
| "loss": 0.339, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.9581320450885669, | |
| "grad_norm": 2.035954816393092, | |
| "learning_rate": 5.334664897442721e-08, | |
| "loss": 0.3376, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 0.9608158883521203, | |
| "grad_norm": 2.045629248133793, | |
| "learning_rate": 4.673687792292436e-08, | |
| "loss": 0.3344, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.9634997316156736, | |
| "grad_norm": 2.073975882418737, | |
| "learning_rate": 4.0562197716448316e-08, | |
| "loss": 0.3314, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 0.966183574879227, | |
| "grad_norm": 1.9890135124600716, | |
| "learning_rate": 3.4823150734186184e-08, | |
| "loss": 0.3355, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.9688674181427804, | |
| "grad_norm": 2.089257154138403, | |
| "learning_rate": 2.9520241089636692e-08, | |
| "loss": 0.3342, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 0.9715512614063339, | |
| "grad_norm": 2.1552227542591122, | |
| "learning_rate": 2.4653934586328365e-08, | |
| "loss": 0.3318, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.9742351046698873, | |
| "grad_norm": 2.0084739911850455, | |
| "learning_rate": 2.022465867690282e-08, | |
| "loss": 0.3264, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 0.9769189479334407, | |
| "grad_norm": 1.935762941530625, | |
| "learning_rate": 1.6232802425568683e-08, | |
| "loss": 0.3377, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.9796027911969941, | |
| "grad_norm": 1.9914145170757653, | |
| "learning_rate": 1.2678716473927821e-08, | |
| "loss": 0.337, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 0.9822866344605475, | |
| "grad_norm": 2.0592560993315936, | |
| "learning_rate": 9.562713010171088e-09, | |
| "loss": 0.3435, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.984970477724101, | |
| "grad_norm": 1.973383618863481, | |
| "learning_rate": 6.885065741661367e-09, | |
| "loss": 0.3407, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 0.9876543209876543, | |
| "grad_norm": 2.0965218988806127, | |
| "learning_rate": 4.646009870885593e-09, | |
| "loss": 0.3477, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.9903381642512077, | |
| "grad_norm": 2.0902332710639597, | |
| "learning_rate": 2.845742074801261e-09, | |
| "loss": 0.3394, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 0.9930220075147611, | |
| "grad_norm": 2.1913038841962558, | |
| "learning_rate": 1.484420487553595e-09, | |
| "loss": 0.3437, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.9957058507783145, | |
| "grad_norm": 2.075178686979665, | |
| "learning_rate": 5.621646865899832e-10, | |
| "loss": 0.3368, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 0.998389694041868, | |
| "grad_norm": 1.997393050323684, | |
| "learning_rate": 7.905568215504921e-11, | |
| "loss": 0.341, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_runtime": 3.3993, | |
| "eval_samples_per_second": 2.942, | |
| "eval_steps_per_second": 0.883, | |
| "step": 1863 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1863, | |
| "total_flos": 195037149265920.0, | |
| "train_loss": 0.0, | |
| "train_runtime": 0.0084, | |
| "train_samples_per_second": 3568486.833, | |
| "train_steps_per_second": 223052.876 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1863, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 195037149265920.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |