| { | |
| "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": 23.707024194155633, | |
| "learning_rate": 5.3475935828877005e-08, | |
| "loss": 1.3383, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0026838432635534087, | |
| "grad_norm": 22.59056007020039, | |
| "learning_rate": 2.6737967914438503e-07, | |
| "loss": 1.3315, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.005367686527106817, | |
| "grad_norm": 13.161641765844873, | |
| "learning_rate": 5.347593582887701e-07, | |
| "loss": 1.2758, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.008051529790660225, | |
| "grad_norm": 12.475140878052374, | |
| "learning_rate": 8.021390374331551e-07, | |
| "loss": 1.1369, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.010735373054213635, | |
| "grad_norm": 6.790566561826281, | |
| "learning_rate": 1.0695187165775401e-06, | |
| "loss": 1.0346, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.013419216317767043, | |
| "grad_norm": 3.5101781959896314, | |
| "learning_rate": 1.3368983957219254e-06, | |
| "loss": 0.9178, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.01610305958132045, | |
| "grad_norm": 3.3276445755680064, | |
| "learning_rate": 1.6042780748663103e-06, | |
| "loss": 0.8689, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.01878690284487386, | |
| "grad_norm": 2.9204742858835586, | |
| "learning_rate": 1.8716577540106954e-06, | |
| "loss": 0.8604, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.02147074610842727, | |
| "grad_norm": 2.90378787155605, | |
| "learning_rate": 2.1390374331550802e-06, | |
| "loss": 0.8468, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.024154589371980676, | |
| "grad_norm": 2.8784414280187463, | |
| "learning_rate": 2.4064171122994653e-06, | |
| "loss": 0.841, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.026838432635534086, | |
| "grad_norm": 2.928562007221605, | |
| "learning_rate": 2.673796791443851e-06, | |
| "loss": 0.8153, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.029522275899087493, | |
| "grad_norm": 2.8761576781715585, | |
| "learning_rate": 2.9411764705882355e-06, | |
| "loss": 0.8119, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.0322061191626409, | |
| "grad_norm": 2.894402518079605, | |
| "learning_rate": 3.2085561497326205e-06, | |
| "loss": 0.8024, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.03488996242619431, | |
| "grad_norm": 2.829747334061854, | |
| "learning_rate": 3.4759358288770056e-06, | |
| "loss": 0.793, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.03757380568974772, | |
| "grad_norm": 2.990925009347966, | |
| "learning_rate": 3.7433155080213907e-06, | |
| "loss": 0.7838, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.040257648953301126, | |
| "grad_norm": 3.120930762964319, | |
| "learning_rate": 4.010695187165775e-06, | |
| "loss": 0.777, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.04294149221685454, | |
| "grad_norm": 3.1035952563975417, | |
| "learning_rate": 4.2780748663101604e-06, | |
| "loss": 0.7574, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.045625335480407946, | |
| "grad_norm": 2.9919150156855103, | |
| "learning_rate": 4.5454545454545455e-06, | |
| "loss": 0.7523, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.04830917874396135, | |
| "grad_norm": 3.0172540493489746, | |
| "learning_rate": 4.812834224598931e-06, | |
| "loss": 0.7464, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.05099302200751476, | |
| "grad_norm": 3.2128136005258643, | |
| "learning_rate": 5.0802139037433165e-06, | |
| "loss": 0.7542, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.05367686527106817, | |
| "grad_norm": 3.2727089445392, | |
| "learning_rate": 5.347593582887702e-06, | |
| "loss": 0.7345, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.05636070853462158, | |
| "grad_norm": 3.194777346407021, | |
| "learning_rate": 5.614973262032086e-06, | |
| "loss": 0.7354, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.059044551798174985, | |
| "grad_norm": 2.9791883890462807, | |
| "learning_rate": 5.882352941176471e-06, | |
| "loss": 0.7407, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.06172839506172839, | |
| "grad_norm": 3.0390205107036303, | |
| "learning_rate": 6.149732620320856e-06, | |
| "loss": 0.7138, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.0644122383252818, | |
| "grad_norm": 3.1517346679317586, | |
| "learning_rate": 6.417112299465241e-06, | |
| "loss": 0.7327, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.06709608158883522, | |
| "grad_norm": 3.04521254750912, | |
| "learning_rate": 6.684491978609626e-06, | |
| "loss": 0.7085, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.06977992485238863, | |
| "grad_norm": 2.9058716602446832, | |
| "learning_rate": 6.951871657754011e-06, | |
| "loss": 0.7208, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.07246376811594203, | |
| "grad_norm": 2.8474752278935282, | |
| "learning_rate": 7.219251336898396e-06, | |
| "loss": 0.7023, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.07514761137949544, | |
| "grad_norm": 2.801020364027186, | |
| "learning_rate": 7.486631016042781e-06, | |
| "loss": 0.7149, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.07783145464304884, | |
| "grad_norm": 2.9190559055219167, | |
| "learning_rate": 7.754010695187166e-06, | |
| "loss": 0.7103, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.08051529790660225, | |
| "grad_norm": 3.2769230507753075, | |
| "learning_rate": 8.02139037433155e-06, | |
| "loss": 0.6976, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.08319914117015566, | |
| "grad_norm": 3.094096427174908, | |
| "learning_rate": 8.288770053475937e-06, | |
| "loss": 0.7008, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.08588298443370908, | |
| "grad_norm": 10.565396166050693, | |
| "learning_rate": 8.556149732620321e-06, | |
| "loss": 0.7168, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.08856682769726248, | |
| "grad_norm": 2.8520403523272937, | |
| "learning_rate": 8.823529411764707e-06, | |
| "loss": 0.7058, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.09125067096081589, | |
| "grad_norm": 2.8358205628742184, | |
| "learning_rate": 9.090909090909091e-06, | |
| "loss": 0.6819, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.0939345142243693, | |
| "grad_norm": 2.805088636732221, | |
| "learning_rate": 9.358288770053477e-06, | |
| "loss": 0.6993, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.0966183574879227, | |
| "grad_norm": 2.7005795024679764, | |
| "learning_rate": 9.625668449197861e-06, | |
| "loss": 0.7029, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.09930220075147611, | |
| "grad_norm": 2.720078515024265, | |
| "learning_rate": 9.893048128342247e-06, | |
| "loss": 0.7118, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.10198604401502952, | |
| "grad_norm": 2.784071908104783, | |
| "learning_rate": 9.999920944317846e-06, | |
| "loss": 0.705, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.10466988727858294, | |
| "grad_norm": 2.7811660208626363, | |
| "learning_rate": 9.99943783531341e-06, | |
| "loss": 0.6941, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.10735373054213634, | |
| "grad_norm": 2.65263132845843, | |
| "learning_rate": 9.998515579512446e-06, | |
| "loss": 0.6978, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.11003757380568975, | |
| "grad_norm": 2.6466962393177758, | |
| "learning_rate": 9.997154257925199e-06, | |
| "loss": 0.7053, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.11272141706924316, | |
| "grad_norm": 2.620753274779537, | |
| "learning_rate": 9.995353990129115e-06, | |
| "loss": 0.6968, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.11540526033279656, | |
| "grad_norm": 2.8859980991251963, | |
| "learning_rate": 9.99311493425834e-06, | |
| "loss": 0.6943, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.11808910359634997, | |
| "grad_norm": 2.6067981225736476, | |
| "learning_rate": 9.99043728698983e-06, | |
| "loss": 0.6905, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.12077294685990338, | |
| "grad_norm": 2.7282158923101654, | |
| "learning_rate": 9.987321283526072e-06, | |
| "loss": 0.6759, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.12345679012345678, | |
| "grad_norm": 2.548636105285145, | |
| "learning_rate": 9.983767197574432e-06, | |
| "loss": 0.6913, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.1261406333870102, | |
| "grad_norm": 2.563304399146852, | |
| "learning_rate": 9.979775341323097e-06, | |
| "loss": 0.682, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.1288244766505636, | |
| "grad_norm": 2.6806528268399905, | |
| "learning_rate": 9.975346065413673e-06, | |
| "loss": 0.6784, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.13150831991411702, | |
| "grad_norm": 2.755997978482572, | |
| "learning_rate": 9.970479758910365e-06, | |
| "loss": 0.6905, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.13419216317767044, | |
| "grad_norm": 2.722046299083315, | |
| "learning_rate": 9.965176849265814e-06, | |
| "loss": 0.6765, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.13687600644122383, | |
| "grad_norm": 2.699702223667157, | |
| "learning_rate": 9.959437802283552e-06, | |
| "loss": 0.6923, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.13955984970477725, | |
| "grad_norm": 15.19444854441991, | |
| "learning_rate": 9.953263122077077e-06, | |
| "loss": 0.6827, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.14224369296833064, | |
| "grad_norm": 2.455056932329254, | |
| "learning_rate": 9.946653351025575e-06, | |
| "loss": 0.6722, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.14492753623188406, | |
| "grad_norm": 2.655065867377216, | |
| "learning_rate": 9.939609069726279e-06, | |
| "loss": 0.6659, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.14761137949543746, | |
| "grad_norm": 2.519302667725412, | |
| "learning_rate": 9.932130896943477e-06, | |
| "loss": 0.6644, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.15029522275899088, | |
| "grad_norm": 2.694996218842278, | |
| "learning_rate": 9.924219489554145e-06, | |
| "loss": 0.6587, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.1529790660225443, | |
| "grad_norm": 2.5115179889905224, | |
| "learning_rate": 9.915875542490257e-06, | |
| "loss": 0.666, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.1556629092860977, | |
| "grad_norm": 2.4116686222381376, | |
| "learning_rate": 9.907099788677745e-06, | |
| "loss": 0.6821, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.1583467525496511, | |
| "grad_norm": 2.5931690654655375, | |
| "learning_rate": 9.897892998972113e-06, | |
| "loss": 0.6634, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.1610305958132045, | |
| "grad_norm": 2.5580006931120796, | |
| "learning_rate": 9.888255982090728e-06, | |
| "loss": 0.6459, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.16371443907675792, | |
| "grad_norm": 2.4769089656127203, | |
| "learning_rate": 9.878189584541783e-06, | |
| "loss": 0.6625, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.16639828234031132, | |
| "grad_norm": 2.5804966552979867, | |
| "learning_rate": 9.867694690549943e-06, | |
| "loss": 0.6608, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.16908212560386474, | |
| "grad_norm": 2.366931727330217, | |
| "learning_rate": 9.85677222197867e-06, | |
| "loss": 0.6714, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.17176596886741816, | |
| "grad_norm": 2.3728211792283207, | |
| "learning_rate": 9.845423138249254e-06, | |
| "loss": 0.6393, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.17444981213097155, | |
| "grad_norm": 2.6556047674716248, | |
| "learning_rate": 9.833648436256525e-06, | |
| "loss": 0.6491, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.17713365539452497, | |
| "grad_norm": 2.394398259028004, | |
| "learning_rate": 9.821449150281308e-06, | |
| "loss": 0.6461, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.17981749865807836, | |
| "grad_norm": 2.4937369392531425, | |
| "learning_rate": 9.808826351899551e-06, | |
| "loss": 0.6565, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.18250134192163178, | |
| "grad_norm": 2.6433202694797697, | |
| "learning_rate": 9.795781149888216e-06, | |
| "loss": 0.643, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.18518518518518517, | |
| "grad_norm": 2.4122988226059077, | |
| "learning_rate": 9.782314690127867e-06, | |
| "loss": 0.644, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.1878690284487386, | |
| "grad_norm": 2.6044944805991435, | |
| "learning_rate": 9.768428155502038e-06, | |
| "loss": 0.6515, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.19055287171229202, | |
| "grad_norm": 2.4666456634619736, | |
| "learning_rate": 9.754122765793306e-06, | |
| "loss": 0.6531, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.1932367149758454, | |
| "grad_norm": 2.6000551658549367, | |
| "learning_rate": 9.739399777576169e-06, | |
| "loss": 0.646, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.19592055823939883, | |
| "grad_norm": 2.4820088229493127, | |
| "learning_rate": 9.72426048410665e-06, | |
| "loss": 0.6363, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.19860440150295222, | |
| "grad_norm": 2.384133235110285, | |
| "learning_rate": 9.70870621520871e-06, | |
| "loss": 0.6325, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.20128824476650564, | |
| "grad_norm": 3.668226993553514, | |
| "learning_rate": 9.692738337157441e-06, | |
| "loss": 0.6401, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.20397208803005903, | |
| "grad_norm": 3.190511998957316, | |
| "learning_rate": 9.676358252559034e-06, | |
| "loss": 0.6423, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.20665593129361245, | |
| "grad_norm": 2.5058637565856543, | |
| "learning_rate": 9.6595674002276e-06, | |
| "loss": 0.6461, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.20933977455716588, | |
| "grad_norm": 2.3544272552545347, | |
| "learning_rate": 9.642367255058767e-06, | |
| "loss": 0.6265, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.21202361782071927, | |
| "grad_norm": 2.3201897189274603, | |
| "learning_rate": 9.624759327900131e-06, | |
| "loss": 0.6289, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.2147074610842727, | |
| "grad_norm": 2.390601051873959, | |
| "learning_rate": 9.606745165418554e-06, | |
| "loss": 0.6417, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.21739130434782608, | |
| "grad_norm": 2.379915435987314, | |
| "learning_rate": 9.58832634996429e-06, | |
| "loss": 0.6284, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.2200751476113795, | |
| "grad_norm": 2.467695540502746, | |
| "learning_rate": 9.569504499432005e-06, | |
| "loss": 0.6165, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.2227589908749329, | |
| "grad_norm": 2.970199159935059, | |
| "learning_rate": 9.550281267118659e-06, | |
| "loss": 0.6411, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.22544283413848631, | |
| "grad_norm": 2.436263404777865, | |
| "learning_rate": 9.530658341578276e-06, | |
| "loss": 0.6339, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.2281266774020397, | |
| "grad_norm": 2.4198695545937636, | |
| "learning_rate": 9.510637446473633e-06, | |
| "loss": 0.6259, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.23081052066559313, | |
| "grad_norm": 2.561893728507687, | |
| "learning_rate": 9.490220340424844e-06, | |
| "loss": 0.6296, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.23349436392914655, | |
| "grad_norm": 2.3758131760337817, | |
| "learning_rate": 9.469408816854898e-06, | |
| "loss": 0.5983, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.23617820719269994, | |
| "grad_norm": 2.533755127048364, | |
| "learning_rate": 9.448204703832102e-06, | |
| "loss": 0.6156, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.23886205045625336, | |
| "grad_norm": 2.3291626236110563, | |
| "learning_rate": 9.426609863909537e-06, | |
| "loss": 0.6103, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.24154589371980675, | |
| "grad_norm": 2.58959045942384, | |
| "learning_rate": 9.404626193961427e-06, | |
| "loss": 0.6183, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.24422973698336017, | |
| "grad_norm": 2.37800172779115, | |
| "learning_rate": 9.382255625016527e-06, | |
| "loss": 0.6204, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.24691358024691357, | |
| "grad_norm": 2.4034236017685577, | |
| "learning_rate": 9.359500122088511e-06, | |
| "loss": 0.611, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.249597423510467, | |
| "grad_norm": 2.5500827608218404, | |
| "learning_rate": 9.336361684003353e-06, | |
| "loss": 0.6119, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.2522812667740204, | |
| "grad_norm": 2.339850893387855, | |
| "learning_rate": 9.312842343223764e-06, | |
| "loss": 0.6139, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.2549651100375738, | |
| "grad_norm": 2.2978994482073385, | |
| "learning_rate": 9.288944165670651e-06, | |
| "loss": 0.6094, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.2576489533011272, | |
| "grad_norm": 2.472531903038008, | |
| "learning_rate": 9.264669250541658e-06, | |
| "loss": 0.5946, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.2603327965646806, | |
| "grad_norm": 2.5419999205504586, | |
| "learning_rate": 9.240019730126764e-06, | |
| "loss": 0.6062, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.26301663982823403, | |
| "grad_norm": 2.500656488650244, | |
| "learning_rate": 9.214997769620998e-06, | |
| "loss": 0.5941, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.26570048309178745, | |
| "grad_norm": 3.0568322611453107, | |
| "learning_rate": 9.189605566934235e-06, | |
| "loss": 0.6033, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.2683843263553409, | |
| "grad_norm": 2.291034606106454, | |
| "learning_rate": 9.163845352498141e-06, | |
| "loss": 0.6033, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.27106816961889424, | |
| "grad_norm": 2.302064075192238, | |
| "learning_rate": 9.137719389070259e-06, | |
| "loss": 0.5964, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.27375201288244766, | |
| "grad_norm": 2.500062438804031, | |
| "learning_rate": 9.111229971535231e-06, | |
| "loss": 0.6136, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.2764358561460011, | |
| "grad_norm": 2.518008812441793, | |
| "learning_rate": 9.084379426703245e-06, | |
| "loss": 0.5889, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.2791196994095545, | |
| "grad_norm": 2.486589468297214, | |
| "learning_rate": 9.057170113105622e-06, | |
| "loss": 0.6058, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.28180354267310787, | |
| "grad_norm": 2.4832636090565416, | |
| "learning_rate": 9.029604420787666e-06, | |
| "loss": 0.5716, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.2844873859366613, | |
| "grad_norm": 2.3087803169454864, | |
| "learning_rate": 9.001684771098709e-06, | |
| "loss": 0.5713, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.2871712292002147, | |
| "grad_norm": 2.4918291692414307, | |
| "learning_rate": 8.973413616479429e-06, | |
| "loss": 0.5862, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.2898550724637681, | |
| "grad_norm": 2.440448292875245, | |
| "learning_rate": 8.944793440246435e-06, | |
| "loss": 0.5729, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.29253891572732155, | |
| "grad_norm": 2.2948744844783517, | |
| "learning_rate": 8.915826756374118e-06, | |
| "loss": 0.5925, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.2952227589908749, | |
| "grad_norm": 2.55433084263396, | |
| "learning_rate": 8.88651610927384e-06, | |
| "loss": 0.5872, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.29790660225442833, | |
| "grad_norm": 2.460150739080302, | |
| "learning_rate": 8.856864073570429e-06, | |
| "loss": 0.5883, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.30059044551798175, | |
| "grad_norm": 2.477680936911416, | |
| "learning_rate": 8.82687325387603e-06, | |
| "loss": 0.5811, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.3032742887815352, | |
| "grad_norm": 2.3268508461918125, | |
| "learning_rate": 8.796546284561307e-06, | |
| "loss": 0.5855, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.3059581320450886, | |
| "grad_norm": 3.8035207505431705, | |
| "learning_rate": 8.765885829524059e-06, | |
| "loss": 0.5771, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.30864197530864196, | |
| "grad_norm": 2.4463363265946727, | |
| "learning_rate": 8.734894581955208e-06, | |
| "loss": 0.5649, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.3113258185721954, | |
| "grad_norm": 2.356445258100388, | |
| "learning_rate": 8.703575264102245e-06, | |
| "loss": 0.5842, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.3140096618357488, | |
| "grad_norm": 2.3797228072331427, | |
| "learning_rate": 8.6719306270301e-06, | |
| "loss": 0.5658, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.3166935050993022, | |
| "grad_norm": 2.5014249915735247, | |
| "learning_rate": 8.639963450379494e-06, | |
| "loss": 0.5606, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.3193773483628556, | |
| "grad_norm": 2.7310240410168114, | |
| "learning_rate": 8.607676542122782e-06, | |
| "loss": 0.5788, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.322061191626409, | |
| "grad_norm": 2.2913621023081876, | |
| "learning_rate": 8.575072738317297e-06, | |
| "loss": 0.5553, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.3247450348899624, | |
| "grad_norm": 2.4074508358318365, | |
| "learning_rate": 8.542154902856232e-06, | |
| "loss": 0.5735, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.32742887815351585, | |
| "grad_norm": 2.504239678740102, | |
| "learning_rate": 8.508925927217083e-06, | |
| "loss": 0.5386, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.33011272141706927, | |
| "grad_norm": 2.2504073190417087, | |
| "learning_rate": 8.475388730207662e-06, | |
| "loss": 0.5467, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.33279656468062263, | |
| "grad_norm": 2.4120408967036084, | |
| "learning_rate": 8.441546257709708e-06, | |
| "loss": 0.5567, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.33548040794417605, | |
| "grad_norm": 2.355093264497884, | |
| "learning_rate": 8.407401482420129e-06, | |
| "loss": 0.557, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.33816425120772947, | |
| "grad_norm": 2.4919190659991988, | |
| "learning_rate": 8.372957403589873e-06, | |
| "loss": 0.578, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.3408480944712829, | |
| "grad_norm": 2.3502885720042137, | |
| "learning_rate": 8.33821704676049e-06, | |
| "loss": 0.5535, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.3435319377348363, | |
| "grad_norm": 2.3670084900296207, | |
| "learning_rate": 8.303183463498357e-06, | |
| "loss": 0.572, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.3462157809983897, | |
| "grad_norm": 2.3219756836196925, | |
| "learning_rate": 8.26785973112664e-06, | |
| "loss": 0.5443, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.3488996242619431, | |
| "grad_norm": 2.436904999678232, | |
| "learning_rate": 8.232248952454978e-06, | |
| "loss": 0.5587, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.3515834675254965, | |
| "grad_norm": 2.3059229586929115, | |
| "learning_rate": 8.196354255506937e-06, | |
| "loss": 0.5612, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.35426731078904994, | |
| "grad_norm": 2.321396440916228, | |
| "learning_rate": 8.160178793245254e-06, | |
| "loss": 0.5527, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.3569511540526033, | |
| "grad_norm": 2.282289091033531, | |
| "learning_rate": 8.12372574329487e-06, | |
| "loss": 0.559, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.3596349973161567, | |
| "grad_norm": 2.344470715222952, | |
| "learning_rate": 8.086998307663815e-06, | |
| "loss": 0.5268, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.36231884057971014, | |
| "grad_norm": 2.3778554642093184, | |
| "learning_rate": 8.049999712461956e-06, | |
| "loss": 0.5478, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.36500268384326356, | |
| "grad_norm": 2.3920121654026705, | |
| "learning_rate": 8.012733207617602e-06, | |
| "loss": 0.5357, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.367686527106817, | |
| "grad_norm": 2.251726458305145, | |
| "learning_rate": 7.975202066592038e-06, | |
| "loss": 0.5355, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.37037037037037035, | |
| "grad_norm": 2.5699454941110575, | |
| "learning_rate": 7.937409586091996e-06, | |
| "loss": 0.5362, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.37305421363392377, | |
| "grad_norm": 2.6392636699570033, | |
| "learning_rate": 7.899359085780062e-06, | |
| "loss": 0.5332, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.3757380568974772, | |
| "grad_norm": 2.4474138908430985, | |
| "learning_rate": 7.86105390798309e-06, | |
| "loss": 0.5316, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.3784219001610306, | |
| "grad_norm": 2.476611610490328, | |
| "learning_rate": 7.822497417398607e-06, | |
| "loss": 0.5374, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.38110574342458403, | |
| "grad_norm": 2.330198719663164, | |
| "learning_rate": 7.783693000799263e-06, | |
| "loss": 0.5364, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.3837895866881374, | |
| "grad_norm": 2.67941726445748, | |
| "learning_rate": 7.744644066735335e-06, | |
| "loss": 0.5547, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.3864734299516908, | |
| "grad_norm": 2.495623767506396, | |
| "learning_rate": 7.705354045235335e-06, | |
| "loss": 0.5225, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.38915727321524424, | |
| "grad_norm": 2.404012336177345, | |
| "learning_rate": 7.665826387504708e-06, | |
| "loss": 0.5335, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.39184111647879766, | |
| "grad_norm": 2.258519467628212, | |
| "learning_rate": 7.626064565622684e-06, | |
| "loss": 0.5352, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.394524959742351, | |
| "grad_norm": 2.302987601546606, | |
| "learning_rate": 7.586072072237291e-06, | |
| "loss": 0.53, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.39720880300590444, | |
| "grad_norm": 2.376869697335436, | |
| "learning_rate": 7.545852420258566e-06, | |
| "loss": 0.5189, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.39989264626945786, | |
| "grad_norm": 2.2721330716946415, | |
| "learning_rate": 7.505409142549987e-06, | |
| "loss": 0.5212, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.4025764895330113, | |
| "grad_norm": 2.3310270375515887, | |
| "learning_rate": 7.464745791618138e-06, | |
| "loss": 0.5246, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.4052603327965647, | |
| "grad_norm": 2.250282725531081, | |
| "learning_rate": 7.423865939300674e-06, | |
| "loss": 0.5406, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.40794417606011807, | |
| "grad_norm": 2.329257450842575, | |
| "learning_rate": 7.382773176452561e-06, | |
| "loss": 0.5279, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.4106280193236715, | |
| "grad_norm": 2.673785473894329, | |
| "learning_rate": 7.341471112630667e-06, | |
| "loss": 0.5142, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.4133118625872249, | |
| "grad_norm": 2.5375662802744636, | |
| "learning_rate": 7.2999633757766956e-06, | |
| "loss": 0.5233, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.41599570585077833, | |
| "grad_norm": 2.3256769903801113, | |
| "learning_rate": 7.258253611898509e-06, | |
| "loss": 0.5136, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.41867954911433175, | |
| "grad_norm": 2.461842880320873, | |
| "learning_rate": 7.216345484749876e-06, | |
| "loss": 0.5263, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.4213633923778851, | |
| "grad_norm": 2.317102515799797, | |
| "learning_rate": 7.174242675508636e-06, | |
| "loss": 0.5152, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.42404723564143854, | |
| "grad_norm": 2.454736200257753, | |
| "learning_rate": 7.131948882453361e-06, | |
| "loss": 0.5117, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.42673107890499196, | |
| "grad_norm": 2.339702221159, | |
| "learning_rate": 7.089467820638491e-06, | |
| "loss": 0.5101, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.4294149221685454, | |
| "grad_norm": 2.3881714482838796, | |
| "learning_rate": 7.0468032215680116e-06, | |
| "loss": 0.5019, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.43209876543209874, | |
| "grad_norm": 2.436454316053831, | |
| "learning_rate": 7.003958832867681e-06, | |
| "loss": 0.5071, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.43478260869565216, | |
| "grad_norm": 2.281577400938406, | |
| "learning_rate": 6.960938417955841e-06, | |
| "loss": 0.5007, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.4374664519592056, | |
| "grad_norm": 2.390481540507048, | |
| "learning_rate": 6.917745755712839e-06, | |
| "loss": 0.5036, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.440150295222759, | |
| "grad_norm": 2.287819577216328, | |
| "learning_rate": 6.874384640149101e-06, | |
| "loss": 0.4975, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.4428341384863124, | |
| "grad_norm": 2.2961340203726044, | |
| "learning_rate": 6.830858880071859e-06, | |
| "loss": 0.5066, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.4455179817498658, | |
| "grad_norm": 2.6654321188888472, | |
| "learning_rate": 6.7871722987505975e-06, | |
| "loss": 0.494, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.4482018250134192, | |
| "grad_norm": 2.309109927856854, | |
| "learning_rate": 6.743328733581211e-06, | |
| "loss": 0.5029, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.45088566827697263, | |
| "grad_norm": 2.3349197671417907, | |
| "learning_rate": 6.6993320357489396e-06, | |
| "loss": 0.5098, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.45356951154052605, | |
| "grad_norm": 2.7102649305338224, | |
| "learning_rate": 6.6551860698900804e-06, | |
| "loss": 0.4938, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.4562533548040794, | |
| "grad_norm": 2.2829956126047586, | |
| "learning_rate": 6.610894713752513e-06, | |
| "loss": 0.4915, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.45893719806763283, | |
| "grad_norm": 2.3074732172577157, | |
| "learning_rate": 6.566461857855096e-06, | |
| "loss": 0.499, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.46162104133118625, | |
| "grad_norm": 2.347122326296506, | |
| "learning_rate": 6.52189140514591e-06, | |
| "loss": 0.4845, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.4643048845947397, | |
| "grad_norm": 2.352714851378591, | |
| "learning_rate": 6.477187270659444e-06, | |
| "loss": 0.4868, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.4669887278582931, | |
| "grad_norm": 2.257212557214909, | |
| "learning_rate": 6.432353381172686e-06, | |
| "loss": 0.4962, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.46967257112184646, | |
| "grad_norm": 2.2555480746952457, | |
| "learning_rate": 6.387393674860205e-06, | |
| "loss": 0.4824, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.4723564143853999, | |
| "grad_norm": 2.15671479307213, | |
| "learning_rate": 6.3423121009482226e-06, | |
| "loss": 0.4917, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.4750402576489533, | |
| "grad_norm": 2.5719647028440265, | |
| "learning_rate": 6.297112619367721e-06, | |
| "loss": 0.4869, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.4777241009125067, | |
| "grad_norm": 2.1577206958571784, | |
| "learning_rate": 6.2517992004066e-06, | |
| "loss": 0.487, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.48040794417606014, | |
| "grad_norm": 2.347019459496274, | |
| "learning_rate": 6.2063758243609275e-06, | |
| "loss": 0.4765, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.4830917874396135, | |
| "grad_norm": 2.484096460788783, | |
| "learning_rate": 6.160846481185326e-06, | |
| "loss": 0.4719, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.4857756307031669, | |
| "grad_norm": 2.294786507809735, | |
| "learning_rate": 6.115215170142481e-06, | |
| "loss": 0.4783, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.48845947396672035, | |
| "grad_norm": 2.275191754224213, | |
| "learning_rate": 6.069485899451863e-06, | |
| "loss": 0.4675, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.49114331723027377, | |
| "grad_norm": 2.259001915851852, | |
| "learning_rate": 6.023662685937643e-06, | |
| "loss": 0.4704, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.49382716049382713, | |
| "grad_norm": 2.3433016730457283, | |
| "learning_rate": 5.9777495546758556e-06, | |
| "loss": 0.4662, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.49651100375738055, | |
| "grad_norm": 2.2115268115453794, | |
| "learning_rate": 5.9317505386408436e-06, | |
| "loss": 0.459, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.499194847020934, | |
| "grad_norm": 2.647545064189506, | |
| "learning_rate": 5.885669678350997e-06, | |
| "loss": 0.4713, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.5018786902844874, | |
| "grad_norm": 2.1992065505187424, | |
| "learning_rate": 5.839511021513853e-06, | |
| "loss": 0.4717, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.5045625335480408, | |
| "grad_norm": 2.3207752638224495, | |
| "learning_rate": 5.793278622670527e-06, | |
| "loss": 0.4643, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.5072463768115942, | |
| "grad_norm": 2.4023558060925367, | |
| "learning_rate": 5.746976542839583e-06, | |
| "loss": 0.4626, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.5099302200751477, | |
| "grad_norm": 2.3325680530570185, | |
| "learning_rate": 5.700608849160307e-06, | |
| "loss": 0.4684, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.512614063338701, | |
| "grad_norm": 2.316854918759411, | |
| "learning_rate": 5.654179614535457e-06, | |
| "loss": 0.4686, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.5152979066022544, | |
| "grad_norm": 2.3326085456533074, | |
| "learning_rate": 5.607692917273499e-06, | |
| "loss": 0.4648, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.5179817498658078, | |
| "grad_norm": 2.289626319103324, | |
| "learning_rate": 5.561152840730371e-06, | |
| "loss": 0.4662, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.5206655931293612, | |
| "grad_norm": 2.3249717568338184, | |
| "learning_rate": 5.514563472950812e-06, | |
| "loss": 0.4593, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.5233494363929146, | |
| "grad_norm": 2.343734137810922, | |
| "learning_rate": 5.46792890630926e-06, | |
| "loss": 0.4596, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.5260332796564681, | |
| "grad_norm": 2.4027428703060085, | |
| "learning_rate": 5.421253237150389e-06, | |
| "loss": 0.4541, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.5287171229200215, | |
| "grad_norm": 2.333576433858016, | |
| "learning_rate": 5.374540565429288e-06, | |
| "loss": 0.4691, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.5314009661835749, | |
| "grad_norm": 2.2510255096964884, | |
| "learning_rate": 5.327794994351321e-06, | |
| "loss": 0.4541, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.5340848094471283, | |
| "grad_norm": 2.492020383298656, | |
| "learning_rate": 5.281020630011703e-06, | |
| "loss": 0.4622, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.5367686527106817, | |
| "grad_norm": 2.2692398941590914, | |
| "learning_rate": 5.234221581034834e-06, | |
| "loss": 0.4432, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.5394524959742351, | |
| "grad_norm": 2.3599001057089226, | |
| "learning_rate": 5.187401958213391e-06, | |
| "loss": 0.4476, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.5421363392377885, | |
| "grad_norm": 2.289539071280456, | |
| "learning_rate": 5.140565874147236e-06, | |
| "loss": 0.4367, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.5448201825013419, | |
| "grad_norm": 2.3731298585226366, | |
| "learning_rate": 5.093717442882185e-06, | |
| "loss": 0.4437, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.5475040257648953, | |
| "grad_norm": 2.442928363781822, | |
| "learning_rate": 5.046860779548613e-06, | |
| "loss": 0.4435, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.5501878690284487, | |
| "grad_norm": 2.2715153478728323, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4362, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.5528717122920022, | |
| "grad_norm": 2.3719270581126546, | |
| "learning_rate": 4.953139220451386e-06, | |
| "loss": 0.4402, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 2.3384700503889566, | |
| "learning_rate": 4.906282557117817e-06, | |
| "loss": 0.4473, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.558239398819109, | |
| "grad_norm": 2.354045727905715, | |
| "learning_rate": 4.859434125852764e-06, | |
| "loss": 0.4488, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.5609232420826624, | |
| "grad_norm": 2.2509560503984023, | |
| "learning_rate": 4.812598041786611e-06, | |
| "loss": 0.4522, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.5636070853462157, | |
| "grad_norm": 2.235469119922816, | |
| "learning_rate": 4.765778418965168e-06, | |
| "loss": 0.4498, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.5662909286097692, | |
| "grad_norm": 2.2464152161649706, | |
| "learning_rate": 4.718979369988299e-06, | |
| "loss": 0.4426, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.5689747718733226, | |
| "grad_norm": 2.294718340510488, | |
| "learning_rate": 4.672205005648681e-06, | |
| "loss": 0.4431, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.571658615136876, | |
| "grad_norm": 2.221625584005981, | |
| "learning_rate": 4.625459434570712e-06, | |
| "loss": 0.4261, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.5743424584004294, | |
| "grad_norm": 2.1935253509326533, | |
| "learning_rate": 4.578746762849612e-06, | |
| "loss": 0.4297, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.5770263016639828, | |
| "grad_norm": 2.299894490322812, | |
| "learning_rate": 4.532071093690741e-06, | |
| "loss": 0.4378, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.5797101449275363, | |
| "grad_norm": 2.2298167009562917, | |
| "learning_rate": 4.485436527049189e-06, | |
| "loss": 0.4204, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.5823939881910897, | |
| "grad_norm": 2.210951499348866, | |
| "learning_rate": 4.438847159269631e-06, | |
| "loss": 0.4376, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.5850778314546431, | |
| "grad_norm": 2.385698410806914, | |
| "learning_rate": 4.392307082726503e-06, | |
| "loss": 0.4227, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.5877616747181964, | |
| "grad_norm": 2.166246773978922, | |
| "learning_rate": 4.345820385464543e-06, | |
| "loss": 0.4167, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.5904455179817498, | |
| "grad_norm": 2.2284183522671146, | |
| "learning_rate": 4.299391150839694e-06, | |
| "loss": 0.424, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.5931293612453032, | |
| "grad_norm": 2.353556576617605, | |
| "learning_rate": 4.253023457160418e-06, | |
| "loss": 0.4263, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.5958132045088567, | |
| "grad_norm": 2.1823241293752567, | |
| "learning_rate": 4.2067213773294745e-06, | |
| "loss": 0.4223, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.5984970477724101, | |
| "grad_norm": 2.3290030938148574, | |
| "learning_rate": 4.16048897848615e-06, | |
| "loss": 0.4123, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.6011808910359635, | |
| "grad_norm": 2.3426832657182035, | |
| "learning_rate": 4.114330321649004e-06, | |
| "loss": 0.4222, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.6038647342995169, | |
| "grad_norm": 2.363754222180504, | |
| "learning_rate": 4.068249461359158e-06, | |
| "loss": 0.4051, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.6065485775630703, | |
| "grad_norm": 2.363465376935289, | |
| "learning_rate": 4.022250445324145e-06, | |
| "loss": 0.4085, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.6092324208266238, | |
| "grad_norm": 2.333761759856766, | |
| "learning_rate": 3.976337314062358e-06, | |
| "loss": 0.4146, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 0.6119162640901772, | |
| "grad_norm": 2.2528760006525435, | |
| "learning_rate": 3.9305141005481385e-06, | |
| "loss": 0.4217, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.6146001073537305, | |
| "grad_norm": 2.2242543084378186, | |
| "learning_rate": 3.8847848298575205e-06, | |
| "loss": 0.4146, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 0.6172839506172839, | |
| "grad_norm": 2.397142677959003, | |
| "learning_rate": 3.839153518814677e-06, | |
| "loss": 0.4076, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.6199677938808373, | |
| "grad_norm": 2.3445507294708157, | |
| "learning_rate": 3.7936241756390746e-06, | |
| "loss": 0.4127, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.6226516371443908, | |
| "grad_norm": 2.3150889117896294, | |
| "learning_rate": 3.748200799593402e-06, | |
| "loss": 0.3881, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.6253354804079442, | |
| "grad_norm": 2.2326861588116844, | |
| "learning_rate": 3.7028873806322797e-06, | |
| "loss": 0.4086, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 0.6280193236714976, | |
| "grad_norm": 2.2298834315348666, | |
| "learning_rate": 3.6576878990517783e-06, | |
| "loss": 0.4043, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.630703166935051, | |
| "grad_norm": 2.1418509986986054, | |
| "learning_rate": 3.6126063251397968e-06, | |
| "loss": 0.4067, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.6333870101986044, | |
| "grad_norm": 2.3237257790530315, | |
| "learning_rate": 3.567646618827315e-06, | |
| "loss": 0.3985, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.6360708534621579, | |
| "grad_norm": 2.3070763336429354, | |
| "learning_rate": 3.522812729340557e-06, | |
| "loss": 0.4011, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 0.6387546967257112, | |
| "grad_norm": 2.2860447287154058, | |
| "learning_rate": 3.4781085948540907e-06, | |
| "loss": 0.4052, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.6414385399892646, | |
| "grad_norm": 2.1248469086865396, | |
| "learning_rate": 3.4335381421449056e-06, | |
| "loss": 0.4035, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 0.644122383252818, | |
| "grad_norm": 2.2786584455625003, | |
| "learning_rate": 3.3891052862474884e-06, | |
| "loss": 0.402, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.6468062265163714, | |
| "grad_norm": 2.502096137168029, | |
| "learning_rate": 3.3448139301099212e-06, | |
| "loss": 0.414, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 0.6494900697799249, | |
| "grad_norm": 2.311720900962173, | |
| "learning_rate": 3.3006679642510596e-06, | |
| "loss": 0.4042, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.6521739130434783, | |
| "grad_norm": 2.237735666858343, | |
| "learning_rate": 3.2566712664187907e-06, | |
| "loss": 0.3913, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 0.6548577563070317, | |
| "grad_norm": 2.182972528696718, | |
| "learning_rate": 3.2128277012494046e-06, | |
| "loss": 0.3977, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.6575415995705851, | |
| "grad_norm": 2.337946486115001, | |
| "learning_rate": 3.169141119928142e-06, | |
| "loss": 0.399, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.6602254428341385, | |
| "grad_norm": 2.2573747340387538, | |
| "learning_rate": 3.125615359850902e-06, | |
| "loss": 0.4006, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.6629092860976918, | |
| "grad_norm": 2.347784470265502, | |
| "learning_rate": 3.082254244287163e-06, | |
| "loss": 0.4114, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 0.6655931293612453, | |
| "grad_norm": 2.1413544102442095, | |
| "learning_rate": 3.0390615820441605e-06, | |
| "loss": 0.4015, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.6682769726247987, | |
| "grad_norm": 2.3307743237747522, | |
| "learning_rate": 2.9960411671323215e-06, | |
| "loss": 0.3939, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 0.6709608158883521, | |
| "grad_norm": 2.262135231401975, | |
| "learning_rate": 2.9531967784319897e-06, | |
| "loss": 0.3999, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.6736446591519055, | |
| "grad_norm": 2.334333363603411, | |
| "learning_rate": 2.9105321793615106e-06, | |
| "loss": 0.4012, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 0.6763285024154589, | |
| "grad_norm": 2.211495094688007, | |
| "learning_rate": 2.8680511175466384e-06, | |
| "loss": 0.3948, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.6790123456790124, | |
| "grad_norm": 2.171871524861147, | |
| "learning_rate": 2.8257573244913637e-06, | |
| "loss": 0.3813, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 0.6816961889425658, | |
| "grad_norm": 2.3365281271089198, | |
| "learning_rate": 2.783654515250126e-06, | |
| "loss": 0.3974, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.6843800322061192, | |
| "grad_norm": 2.2433054343843053, | |
| "learning_rate": 2.741746388101493e-06, | |
| "loss": 0.3845, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.6870638754696726, | |
| "grad_norm": 2.15211707144342, | |
| "learning_rate": 2.700036624223308e-06, | |
| "loss": 0.3918, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.6897477187332259, | |
| "grad_norm": 2.1996998018685154, | |
| "learning_rate": 2.6585288873693355e-06, | |
| "loss": 0.398, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 0.6924315619967794, | |
| "grad_norm": 2.106192506014751, | |
| "learning_rate": 2.6172268235474396e-06, | |
| "loss": 0.3951, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.6951154052603328, | |
| "grad_norm": 2.238179206766993, | |
| "learning_rate": 2.576134060699328e-06, | |
| "loss": 0.3924, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 0.6977992485238862, | |
| "grad_norm": 2.1297812983505637, | |
| "learning_rate": 2.5352542083818636e-06, | |
| "loss": 0.382, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.7004830917874396, | |
| "grad_norm": 2.3189960201182958, | |
| "learning_rate": 2.494590857450016e-06, | |
| "loss": 0.3875, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 0.703166935050993, | |
| "grad_norm": 2.264897141770745, | |
| "learning_rate": 2.454147579741436e-06, | |
| "loss": 0.3808, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.7058507783145465, | |
| "grad_norm": 2.223118660587265, | |
| "learning_rate": 2.4139279277627113e-06, | |
| "loss": 0.3927, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 0.7085346215780999, | |
| "grad_norm": 2.284889395217115, | |
| "learning_rate": 2.373935434377316e-06, | |
| "loss": 0.3852, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.7112184648416533, | |
| "grad_norm": 2.2020638584782124, | |
| "learning_rate": 2.3341736124952924e-06, | |
| "loss": 0.3759, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.7139023081052066, | |
| "grad_norm": 2.112679516796632, | |
| "learning_rate": 2.294645954764666e-06, | |
| "loss": 0.3898, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.71658615136876, | |
| "grad_norm": 2.2435774581489825, | |
| "learning_rate": 2.2553559332646675e-06, | |
| "loss": 0.3726, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 0.7192699946323134, | |
| "grad_norm": 2.1884891997535294, | |
| "learning_rate": 2.216306999200739e-06, | |
| "loss": 0.377, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.7219538378958669, | |
| "grad_norm": 2.294647098985639, | |
| "learning_rate": 2.1775025826013953e-06, | |
| "loss": 0.3758, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 0.7246376811594203, | |
| "grad_norm": 2.22553650274205, | |
| "learning_rate": 2.1389460920169105e-06, | |
| "loss": 0.3817, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.7273215244229737, | |
| "grad_norm": 2.1609674662857015, | |
| "learning_rate": 2.100640914219939e-06, | |
| "loss": 0.3862, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 0.7300053676865271, | |
| "grad_norm": 2.1197685045819004, | |
| "learning_rate": 2.062590413908007e-06, | |
| "loss": 0.3691, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.7326892109500805, | |
| "grad_norm": 2.1688165816422367, | |
| "learning_rate": 2.0247979334079646e-06, | |
| "loss": 0.3857, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 0.735373054213634, | |
| "grad_norm": 2.229655685961629, | |
| "learning_rate": 1.9872667923823996e-06, | |
| "loss": 0.3687, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.7380568974771873, | |
| "grad_norm": 2.1365838274814135, | |
| "learning_rate": 1.9500002875380458e-06, | |
| "loss": 0.3604, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "grad_norm": 2.242861382719449, | |
| "learning_rate": 1.9130016923361843e-06, | |
| "loss": 0.3801, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.7434245840042941, | |
| "grad_norm": 2.3006806690999246, | |
| "learning_rate": 1.8762742567051318e-06, | |
| "loss": 0.3694, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 0.7461084272678475, | |
| "grad_norm": 2.218533870827561, | |
| "learning_rate": 1.8398212067547455e-06, | |
| "loss": 0.3749, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.748792270531401, | |
| "grad_norm": 2.211081714806873, | |
| "learning_rate": 1.8036457444930643e-06, | |
| "loss": 0.3675, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 0.7514761137949544, | |
| "grad_norm": 2.274070404317869, | |
| "learning_rate": 1.7677510475450244e-06, | |
| "loss": 0.3682, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.7541599570585078, | |
| "grad_norm": 2.1717478115680895, | |
| "learning_rate": 1.7321402688733618e-06, | |
| "loss": 0.3639, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 0.7568438003220612, | |
| "grad_norm": 2.090172688253105, | |
| "learning_rate": 1.6968165365016443e-06, | |
| "loss": 0.3671, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.7595276435856146, | |
| "grad_norm": 2.152111966043745, | |
| "learning_rate": 1.661782953239512e-06, | |
| "loss": 0.3681, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 0.7622114868491681, | |
| "grad_norm": 2.1912001313892495, | |
| "learning_rate": 1.6270425964101272e-06, | |
| "loss": 0.3762, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.7648953301127214, | |
| "grad_norm": 2.161939449796454, | |
| "learning_rate": 1.5925985175798741e-06, | |
| "loss": 0.3607, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.7675791733762748, | |
| "grad_norm": 2.1320980119174138, | |
| "learning_rate": 1.5584537422902923e-06, | |
| "loss": 0.3646, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.7702630166398282, | |
| "grad_norm": 2.0239291711448635, | |
| "learning_rate": 1.5246112697923389e-06, | |
| "loss": 0.3567, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 0.7729468599033816, | |
| "grad_norm": 2.2747263764152486, | |
| "learning_rate": 1.4910740727829176e-06, | |
| "loss": 0.3605, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.775630703166935, | |
| "grad_norm": 2.0481307516954956, | |
| "learning_rate": 1.457845097143769e-06, | |
| "loss": 0.3542, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 0.7783145464304885, | |
| "grad_norm": 2.150697775737251, | |
| "learning_rate": 1.424927261682703e-06, | |
| "loss": 0.3576, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.7809983896940419, | |
| "grad_norm": 2.0211316558629653, | |
| "learning_rate": 1.3923234578772177e-06, | |
| "loss": 0.3589, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 0.7836822329575953, | |
| "grad_norm": 1.9914896430875126, | |
| "learning_rate": 1.3600365496205059e-06, | |
| "loss": 0.362, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.7863660762211487, | |
| "grad_norm": 2.04982934560283, | |
| "learning_rate": 1.3280693729699018e-06, | |
| "loss": 0.3483, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 0.789049919484702, | |
| "grad_norm": 2.151808291733713, | |
| "learning_rate": 1.2964247358977567e-06, | |
| "loss": 0.3528, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.7917337627482555, | |
| "grad_norm": 2.182330920063843, | |
| "learning_rate": 1.265105418044793e-06, | |
| "loss": 0.3528, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.7944176060118089, | |
| "grad_norm": 2.054495838591551, | |
| "learning_rate": 1.2341141704759418e-06, | |
| "loss": 0.37, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.7971014492753623, | |
| "grad_norm": 2.108984513161976, | |
| "learning_rate": 1.2034537154386933e-06, | |
| "loss": 0.3499, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 0.7997852925389157, | |
| "grad_norm": 2.125942404071833, | |
| "learning_rate": 1.173126746123972e-06, | |
| "loss": 0.3623, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.8024691358024691, | |
| "grad_norm": 1.9949859699489827, | |
| "learning_rate": 1.1431359264295717e-06, | |
| "loss": 0.356, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 0.8051529790660226, | |
| "grad_norm": 2.0802922279223646, | |
| "learning_rate": 1.113483890726162e-06, | |
| "loss": 0.3538, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.807836822329576, | |
| "grad_norm": 1.9638277128766195, | |
| "learning_rate": 1.084173243625884e-06, | |
| "loss": 0.3444, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 0.8105206655931294, | |
| "grad_norm": 2.5501494543487007, | |
| "learning_rate": 1.0552065597535671e-06, | |
| "loss": 0.3601, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.8132045088566827, | |
| "grad_norm": 2.083272873220911, | |
| "learning_rate": 1.0265863835205709e-06, | |
| "loss": 0.352, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 0.8158883521202361, | |
| "grad_norm": 2.175472987287477, | |
| "learning_rate": 9.983152289012926e-07, | |
| "loss": 0.3591, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.8185721953837896, | |
| "grad_norm": 2.1999371974208892, | |
| "learning_rate": 9.70395579212336e-07, | |
| "loss": 0.3536, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.821256038647343, | |
| "grad_norm": 2.031297723323445, | |
| "learning_rate": 9.428298868943791e-07, | |
| "loss": 0.3511, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.8239398819108964, | |
| "grad_norm": 2.1091091385698957, | |
| "learning_rate": 9.15620573296756e-07, | |
| "loss": 0.3434, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 0.8266237251744498, | |
| "grad_norm": 2.252603598055895, | |
| "learning_rate": 8.887700284647699e-07, | |
| "loss": 0.3582, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.8293075684380032, | |
| "grad_norm": 1.9675118256158026, | |
| "learning_rate": 8.622806109297432e-07, | |
| "loss": 0.3451, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 0.8319914117015567, | |
| "grad_norm": 2.1525535225266568, | |
| "learning_rate": 8.361546475018589e-07, | |
| "loss": 0.3636, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.8346752549651101, | |
| "grad_norm": 2.2811796548762207, | |
| "learning_rate": 8.103944330657665e-07, | |
| "loss": 0.3483, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 0.8373590982286635, | |
| "grad_norm": 2.0199594885732455, | |
| "learning_rate": 7.850022303790033e-07, | |
| "loss": 0.3413, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.8400429414922168, | |
| "grad_norm": 2.2070824777860767, | |
| "learning_rate": 7.599802698732356e-07, | |
| "loss": 0.3512, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 0.8427267847557702, | |
| "grad_norm": 2.122267143101352, | |
| "learning_rate": 7.353307494583445e-07, | |
| "loss": 0.3538, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.8454106280193237, | |
| "grad_norm": 2.032897527740881, | |
| "learning_rate": 7.1105583432935e-07, | |
| "loss": 0.3517, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 0.8480944712828771, | |
| "grad_norm": 2.3012343967020596, | |
| "learning_rate": 6.871576567762373e-07, | |
| "loss": 0.3549, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.8507783145464305, | |
| "grad_norm": 2.0359890432050185, | |
| "learning_rate": 6.63638315996647e-07, | |
| "loss": 0.341, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 0.8534621578099839, | |
| "grad_norm": 1.9376850713401, | |
| "learning_rate": 6.404998779114912e-07, | |
| "loss": 0.3429, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.8561460010735373, | |
| "grad_norm": 2.0142712450924125, | |
| "learning_rate": 6.177443749834743e-07, | |
| "loss": 0.3354, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 0.8588298443370908, | |
| "grad_norm": 2.3456418462427373, | |
| "learning_rate": 5.953738060385761e-07, | |
| "loss": 0.3569, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.8615136876006442, | |
| "grad_norm": 2.159644228977299, | |
| "learning_rate": 5.733901360904648e-07, | |
| "loss": 0.348, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 0.8641975308641975, | |
| "grad_norm": 2.0418625012147213, | |
| "learning_rate": 5.517952961678997e-07, | |
| "loss": 0.3481, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.8668813741277509, | |
| "grad_norm": 1.9221731825522803, | |
| "learning_rate": 5.305911831451044e-07, | |
| "loss": 0.3386, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "grad_norm": 2.1099798966385492, | |
| "learning_rate": 5.097796595751564e-07, | |
| "loss": 0.3492, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.8722490606548577, | |
| "grad_norm": 1.99915346290055, | |
| "learning_rate": 4.89362553526368e-07, | |
| "loss": 0.3444, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 0.8749329039184112, | |
| "grad_norm": 2.0762456669317015, | |
| "learning_rate": 4.6934165842172476e-07, | |
| "loss": 0.3516, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.8776167471819646, | |
| "grad_norm": 2.1492424701034416, | |
| "learning_rate": 4.4971873288134237e-07, | |
| "loss": 0.3336, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 0.880300590445518, | |
| "grad_norm": 2.0290558303807913, | |
| "learning_rate": 4.304955005679962e-07, | |
| "loss": 0.3496, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.8829844337090714, | |
| "grad_norm": 1.9971182002696468, | |
| "learning_rate": 4.1167365003571047e-07, | |
| "loss": 0.3404, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 0.8856682769726248, | |
| "grad_norm": 2.032567293276667, | |
| "learning_rate": 3.9325483458144684e-07, | |
| "loss": 0.347, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.8883521202361782, | |
| "grad_norm": 2.1850140352772844, | |
| "learning_rate": 3.752406720998691e-07, | |
| "loss": 0.3445, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 0.8910359634997316, | |
| "grad_norm": 2.011004156429553, | |
| "learning_rate": 3.5763274494123424e-07, | |
| "loss": 0.3315, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.893719806763285, | |
| "grad_norm": 1.9859631872738064, | |
| "learning_rate": 3.404325997723995e-07, | |
| "loss": 0.3344, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 0.8964036500268384, | |
| "grad_norm": 2.1074476847700754, | |
| "learning_rate": 3.2364174744096587e-07, | |
| "loss": 0.3501, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.8990874932903918, | |
| "grad_norm": 2.197381464597478, | |
| "learning_rate": 3.072616628425601e-07, | |
| "loss": 0.3438, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 0.9017713365539453, | |
| "grad_norm": 1.9701476428536195, | |
| "learning_rate": 2.9129378479129e-07, | |
| "loss": 0.3346, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.9044551798174987, | |
| "grad_norm": 1.905205704349654, | |
| "learning_rate": 2.7573951589335177e-07, | |
| "loss": 0.3401, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 0.9071390230810521, | |
| "grad_norm": 1.9895541805632337, | |
| "learning_rate": 2.60600222423833e-07, | |
| "loss": 0.3382, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.9098228663446055, | |
| "grad_norm": 2.0417279425409833, | |
| "learning_rate": 2.45877234206694e-07, | |
| "loss": 0.3389, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 0.9125067096081588, | |
| "grad_norm": 1.929413840069086, | |
| "learning_rate": 2.3157184449796365e-07, | |
| "loss": 0.3395, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.9151905528717122, | |
| "grad_norm": 2.083117283846843, | |
| "learning_rate": 2.1768530987213267e-07, | |
| "loss": 0.3403, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 0.9178743961352657, | |
| "grad_norm": 2.0593648248391903, | |
| "learning_rate": 2.0421885011178532e-07, | |
| "loss": 0.3277, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.9205582393988191, | |
| "grad_norm": 2.0026179811647684, | |
| "learning_rate": 1.911736481004489e-07, | |
| "loss": 0.3185, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 0.9232420826623725, | |
| "grad_norm": 2.1022822626142057, | |
| "learning_rate": 1.7855084971869385e-07, | |
| "loss": 0.3465, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.9259259259259259, | |
| "grad_norm": 1.9614578906671158, | |
| "learning_rate": 1.663515637434765e-07, | |
| "loss": 0.3417, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 0.9286097691894794, | |
| "grad_norm": 1.9913289680582344, | |
| "learning_rate": 1.5457686175074816e-07, | |
| "loss": 0.3344, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.9312936124530328, | |
| "grad_norm": 1.990691441342211, | |
| "learning_rate": 1.432277780213298e-07, | |
| "loss": 0.3323, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 0.9339774557165862, | |
| "grad_norm": 2.13052260029459, | |
| "learning_rate": 1.3230530945005704e-07, | |
| "loss": 0.3311, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.9366612989801396, | |
| "grad_norm": 1.9831089635128403, | |
| "learning_rate": 1.2181041545821727e-07, | |
| "loss": 0.3402, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 0.9393451422436929, | |
| "grad_norm": 2.0550140670951977, | |
| "learning_rate": 1.1174401790927325e-07, | |
| "loss": 0.3376, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.9420289855072463, | |
| "grad_norm": 2.102896894928286, | |
| "learning_rate": 1.0210700102788796e-07, | |
| "loss": 0.3405, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 0.9447128287707998, | |
| "grad_norm": 2.0386910430067955, | |
| "learning_rate": 9.290021132225546e-08, | |
| "loss": 0.3335, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.9473966720343532, | |
| "grad_norm": 1.9753486047261384, | |
| "learning_rate": 8.412445750974385e-08, | |
| "loss": 0.3331, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 0.9500805152979066, | |
| "grad_norm": 2.0982536596669363, | |
| "learning_rate": 7.578051044585644e-08, | |
| "loss": 0.3304, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.95276435856146, | |
| "grad_norm": 1.883254499694507, | |
| "learning_rate": 6.786910305652373e-08, | |
| "loss": 0.3328, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 0.9554482018250134, | |
| "grad_norm": 2.000281090341156, | |
| "learning_rate": 6.0390930273721e-08, | |
| "loss": 0.3453, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.9581320450885669, | |
| "grad_norm": 2.143507539729109, | |
| "learning_rate": 5.334664897442721e-08, | |
| "loss": 0.3531, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 0.9608158883521203, | |
| "grad_norm": 2.0784537700051184, | |
| "learning_rate": 4.673687792292436e-08, | |
| "loss": 0.3335, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.9634997316156736, | |
| "grad_norm": 1.9630462834561795, | |
| "learning_rate": 4.0562197716448316e-08, | |
| "loss": 0.3359, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 0.966183574879227, | |
| "grad_norm": 1.958212519833783, | |
| "learning_rate": 3.4823150734186184e-08, | |
| "loss": 0.3322, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.9688674181427804, | |
| "grad_norm": 2.1550621294997963, | |
| "learning_rate": 2.9520241089636692e-08, | |
| "loss": 0.3319, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 0.9715512614063339, | |
| "grad_norm": 2.2101816114526063, | |
| "learning_rate": 2.4653934586328365e-08, | |
| "loss": 0.3373, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.9742351046698873, | |
| "grad_norm": 2.2419490107739324, | |
| "learning_rate": 2.022465867690282e-08, | |
| "loss": 0.3433, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 0.9769189479334407, | |
| "grad_norm": 2.1034628009481358, | |
| "learning_rate": 1.6232802425568683e-08, | |
| "loss": 0.3386, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.9796027911969941, | |
| "grad_norm": 2.0401577654553287, | |
| "learning_rate": 1.2678716473927821e-08, | |
| "loss": 0.3413, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 0.9822866344605475, | |
| "grad_norm": 2.225206626746121, | |
| "learning_rate": 9.562713010171088e-09, | |
| "loss": 0.3348, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.984970477724101, | |
| "grad_norm": 2.124211075922287, | |
| "learning_rate": 6.885065741661367e-09, | |
| "loss": 0.3493, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 0.9876543209876543, | |
| "grad_norm": 2.1755251808066043, | |
| "learning_rate": 4.646009870885593e-09, | |
| "loss": 0.3362, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.9903381642512077, | |
| "grad_norm": 1.9898618189971538, | |
| "learning_rate": 2.845742074801261e-09, | |
| "loss": 0.3369, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 0.9930220075147611, | |
| "grad_norm": 2.1164231384263754, | |
| "learning_rate": 1.484420487553595e-09, | |
| "loss": 0.3307, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.9957058507783145, | |
| "grad_norm": 1.9456813410292053, | |
| "learning_rate": 5.621646865899832e-10, | |
| "loss": 0.336, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 0.998389694041868, | |
| "grad_norm": 2.1400926686598445, | |
| "learning_rate": 7.905568215504921e-11, | |
| "loss": 0.3342, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_runtime": 3.3032, | |
| "eval_samples_per_second": 3.027, | |
| "eval_steps_per_second": 0.908, | |
| "step": 1863 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1863, | |
| "total_flos": 195037149265920.0, | |
| "train_loss": 0.5027870363403377, | |
| "train_runtime": 16520.2246, | |
| "train_samples_per_second": 1.804, | |
| "train_steps_per_second": 0.113 | |
| } | |
| ], | |
| "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 | |
| } | |