| { | |
| "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": 23.28716171727923, | |
| "learning_rate": 5.3475935828877005e-08, | |
| "loss": 1.3606, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0026838432635534087, | |
| "grad_norm": 22.769137088546284, | |
| "learning_rate": 2.6737967914438503e-07, | |
| "loss": 1.3206, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.005367686527106817, | |
| "grad_norm": 15.03004000090198, | |
| "learning_rate": 5.347593582887701e-07, | |
| "loss": 1.2762, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.008051529790660225, | |
| "grad_norm": 12.378775400312193, | |
| "learning_rate": 8.021390374331551e-07, | |
| "loss": 1.1378, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.010735373054213635, | |
| "grad_norm": 8.125374906522033, | |
| "learning_rate": 1.0695187165775401e-06, | |
| "loss": 1.0346, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.013419216317767043, | |
| "grad_norm": 3.5925170353696565, | |
| "learning_rate": 1.3368983957219254e-06, | |
| "loss": 0.9282, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.01610305958132045, | |
| "grad_norm": 3.4343761488143065, | |
| "learning_rate": 1.6042780748663103e-06, | |
| "loss": 0.8844, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.01878690284487386, | |
| "grad_norm": 2.9595304178432196, | |
| "learning_rate": 1.8716577540106954e-06, | |
| "loss": 0.8636, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.02147074610842727, | |
| "grad_norm": 2.9399370698169567, | |
| "learning_rate": 2.1390374331550802e-06, | |
| "loss": 0.8473, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.024154589371980676, | |
| "grad_norm": 2.9713501019978614, | |
| "learning_rate": 2.4064171122994653e-06, | |
| "loss": 0.8181, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.026838432635534086, | |
| "grad_norm": 3.0584787370054074, | |
| "learning_rate": 2.673796791443851e-06, | |
| "loss": 0.8237, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.029522275899087493, | |
| "grad_norm": 3.017155094320408, | |
| "learning_rate": 2.9411764705882355e-06, | |
| "loss": 0.8148, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.0322061191626409, | |
| "grad_norm": 3.1200001832710593, | |
| "learning_rate": 3.2085561497326205e-06, | |
| "loss": 0.7871, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.03488996242619431, | |
| "grad_norm": 3.044693130677567, | |
| "learning_rate": 3.4759358288770056e-06, | |
| "loss": 0.7864, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.03757380568974772, | |
| "grad_norm": 3.122076906807405, | |
| "learning_rate": 3.7433155080213907e-06, | |
| "loss": 0.7698, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.040257648953301126, | |
| "grad_norm": 3.1183701395947843, | |
| "learning_rate": 4.010695187165775e-06, | |
| "loss": 0.7654, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.04294149221685454, | |
| "grad_norm": 3.3424940821294693, | |
| "learning_rate": 4.2780748663101604e-06, | |
| "loss": 0.7674, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.045625335480407946, | |
| "grad_norm": 3.2035321937108883, | |
| "learning_rate": 4.5454545454545455e-06, | |
| "loss": 0.7499, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.04830917874396135, | |
| "grad_norm": 3.5212471784692534, | |
| "learning_rate": 4.812834224598931e-06, | |
| "loss": 0.7643, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.05099302200751476, | |
| "grad_norm": 3.2363611834200783, | |
| "learning_rate": 5.0802139037433165e-06, | |
| "loss": 0.7313, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.05367686527106817, | |
| "grad_norm": 3.073834880911029, | |
| "learning_rate": 5.347593582887702e-06, | |
| "loss": 0.7344, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.05636070853462158, | |
| "grad_norm": 3.141450665557805, | |
| "learning_rate": 5.614973262032086e-06, | |
| "loss": 0.7254, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.059044551798174985, | |
| "grad_norm": 3.0516521468853375, | |
| "learning_rate": 5.882352941176471e-06, | |
| "loss": 0.7244, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.06172839506172839, | |
| "grad_norm": 3.178437501326201, | |
| "learning_rate": 6.149732620320856e-06, | |
| "loss": 0.7367, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.0644122383252818, | |
| "grad_norm": 2.899640273081663, | |
| "learning_rate": 6.417112299465241e-06, | |
| "loss": 0.7159, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.06709608158883522, | |
| "grad_norm": 3.0078658981215263, | |
| "learning_rate": 6.684491978609626e-06, | |
| "loss": 0.7143, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.06977992485238863, | |
| "grad_norm": 2.920930952013271, | |
| "learning_rate": 6.951871657754011e-06, | |
| "loss": 0.708, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.07246376811594203, | |
| "grad_norm": 2.9407712070114775, | |
| "learning_rate": 7.219251336898396e-06, | |
| "loss": 0.7173, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.07514761137949544, | |
| "grad_norm": 3.0787680794058825, | |
| "learning_rate": 7.486631016042781e-06, | |
| "loss": 0.716, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.07783145464304884, | |
| "grad_norm": 2.854092199561474, | |
| "learning_rate": 7.754010695187166e-06, | |
| "loss": 0.705, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.08051529790660225, | |
| "grad_norm": 2.8725262562217733, | |
| "learning_rate": 8.02139037433155e-06, | |
| "loss": 0.6948, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.08319914117015566, | |
| "grad_norm": 2.9188370284915006, | |
| "learning_rate": 8.288770053475937e-06, | |
| "loss": 0.7071, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.08588298443370908, | |
| "grad_norm": 2.9837446626262785, | |
| "learning_rate": 8.556149732620321e-06, | |
| "loss": 0.7065, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.08856682769726248, | |
| "grad_norm": 2.8792102185639568, | |
| "learning_rate": 8.823529411764707e-06, | |
| "loss": 0.7189, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.09125067096081589, | |
| "grad_norm": 2.9038223427057828, | |
| "learning_rate": 9.090909090909091e-06, | |
| "loss": 0.7049, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.0939345142243693, | |
| "grad_norm": 2.794069021471311, | |
| "learning_rate": 9.358288770053477e-06, | |
| "loss": 0.7132, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.0966183574879227, | |
| "grad_norm": 2.7067368920123, | |
| "learning_rate": 9.625668449197861e-06, | |
| "loss": 0.6938, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.09930220075147611, | |
| "grad_norm": 2.675585819188982, | |
| "learning_rate": 9.893048128342247e-06, | |
| "loss": 0.7122, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.10198604401502952, | |
| "grad_norm": 2.7228236046059116, | |
| "learning_rate": 9.999920944317846e-06, | |
| "loss": 0.6984, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.10466988727858294, | |
| "grad_norm": 2.7541945623220725, | |
| "learning_rate": 9.99943783531341e-06, | |
| "loss": 0.6952, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.10735373054213634, | |
| "grad_norm": 2.7994365909423906, | |
| "learning_rate": 9.998515579512446e-06, | |
| "loss": 0.7039, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.11003757380568975, | |
| "grad_norm": 2.598440628370381, | |
| "learning_rate": 9.997154257925199e-06, | |
| "loss": 0.7026, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.11272141706924316, | |
| "grad_norm": 2.687502171080451, | |
| "learning_rate": 9.995353990129115e-06, | |
| "loss": 0.6916, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.11540526033279656, | |
| "grad_norm": 2.742627030622923, | |
| "learning_rate": 9.99311493425834e-06, | |
| "loss": 0.6911, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.11808910359634997, | |
| "grad_norm": 2.707206937929519, | |
| "learning_rate": 9.99043728698983e-06, | |
| "loss": 0.6865, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.12077294685990338, | |
| "grad_norm": 2.78586456787663, | |
| "learning_rate": 9.987321283526072e-06, | |
| "loss": 0.6781, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.12345679012345678, | |
| "grad_norm": 2.609935225307783, | |
| "learning_rate": 9.983767197574432e-06, | |
| "loss": 0.6827, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.1261406333870102, | |
| "grad_norm": 3.8321019619689163, | |
| "learning_rate": 9.979775341323097e-06, | |
| "loss": 0.6906, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.1288244766505636, | |
| "grad_norm": 2.6458670648252056, | |
| "learning_rate": 9.975346065413673e-06, | |
| "loss": 0.6715, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.13150831991411702, | |
| "grad_norm": 2.577401888084613, | |
| "learning_rate": 9.970479758910365e-06, | |
| "loss": 0.6748, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.13419216317767044, | |
| "grad_norm": 2.8481385954442864, | |
| "learning_rate": 9.965176849265814e-06, | |
| "loss": 0.6754, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.13687600644122383, | |
| "grad_norm": 2.517246019195469, | |
| "learning_rate": 9.959437802283552e-06, | |
| "loss": 0.682, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.13955984970477725, | |
| "grad_norm": 2.6167930050900012, | |
| "learning_rate": 9.953263122077077e-06, | |
| "loss": 0.6679, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.14224369296833064, | |
| "grad_norm": 2.6780218273512832, | |
| "learning_rate": 9.946653351025575e-06, | |
| "loss": 0.6904, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.14492753623188406, | |
| "grad_norm": 2.5393008267043244, | |
| "learning_rate": 9.939609069726279e-06, | |
| "loss": 0.6588, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.14761137949543746, | |
| "grad_norm": 2.567260216330596, | |
| "learning_rate": 9.932130896943477e-06, | |
| "loss": 0.6764, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.15029522275899088, | |
| "grad_norm": 2.6695824974327538, | |
| "learning_rate": 9.924219489554145e-06, | |
| "loss": 0.6813, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.1529790660225443, | |
| "grad_norm": 2.533940793717001, | |
| "learning_rate": 9.915875542490257e-06, | |
| "loss": 0.6536, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.1556629092860977, | |
| "grad_norm": 2.5674293408462936, | |
| "learning_rate": 9.907099788677745e-06, | |
| "loss": 0.6559, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.1583467525496511, | |
| "grad_norm": 2.8337795609609975, | |
| "learning_rate": 9.897892998972113e-06, | |
| "loss": 0.6661, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.1610305958132045, | |
| "grad_norm": 2.3297630789068586, | |
| "learning_rate": 9.888255982090728e-06, | |
| "loss": 0.6531, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.16371443907675792, | |
| "grad_norm": 2.4718746964525957, | |
| "learning_rate": 9.878189584541783e-06, | |
| "loss": 0.6587, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.16639828234031132, | |
| "grad_norm": 2.499824010266399, | |
| "learning_rate": 9.867694690549943e-06, | |
| "loss": 0.6631, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.16908212560386474, | |
| "grad_norm": 2.6049840516161176, | |
| "learning_rate": 9.85677222197867e-06, | |
| "loss": 0.6703, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.17176596886741816, | |
| "grad_norm": 2.5299411616320433, | |
| "learning_rate": 9.845423138249254e-06, | |
| "loss": 0.6542, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.17444981213097155, | |
| "grad_norm": 2.616452811741115, | |
| "learning_rate": 9.833648436256525e-06, | |
| "loss": 0.6474, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.17713365539452497, | |
| "grad_norm": 2.961763150997931, | |
| "learning_rate": 9.821449150281308e-06, | |
| "loss": 0.6523, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.17981749865807836, | |
| "grad_norm": 2.3905734159429484, | |
| "learning_rate": 9.808826351899551e-06, | |
| "loss": 0.6513, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.18250134192163178, | |
| "grad_norm": 2.3984015098146445, | |
| "learning_rate": 9.795781149888216e-06, | |
| "loss": 0.6545, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.18518518518518517, | |
| "grad_norm": 2.5747475475768997, | |
| "learning_rate": 9.782314690127867e-06, | |
| "loss": 0.6503, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.1878690284487386, | |
| "grad_norm": 2.5437744632970305, | |
| "learning_rate": 9.768428155502038e-06, | |
| "loss": 0.6368, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.19055287171229202, | |
| "grad_norm": 2.4483397168998344, | |
| "learning_rate": 9.754122765793306e-06, | |
| "loss": 0.651, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.1932367149758454, | |
| "grad_norm": 2.41502899586079, | |
| "learning_rate": 9.739399777576169e-06, | |
| "loss": 0.6388, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.19592055823939883, | |
| "grad_norm": 2.4453852895114165, | |
| "learning_rate": 9.72426048410665e-06, | |
| "loss": 0.6473, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.19860440150295222, | |
| "grad_norm": 2.6224111914099817, | |
| "learning_rate": 9.70870621520871e-06, | |
| "loss": 0.6391, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.20128824476650564, | |
| "grad_norm": 2.4204256543979823, | |
| "learning_rate": 9.692738337157441e-06, | |
| "loss": 0.6344, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.20397208803005903, | |
| "grad_norm": 2.507545837245254, | |
| "learning_rate": 9.676358252559034e-06, | |
| "loss": 0.6442, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.20665593129361245, | |
| "grad_norm": 2.447395496493652, | |
| "learning_rate": 9.6595674002276e-06, | |
| "loss": 0.64, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.20933977455716588, | |
| "grad_norm": 2.535000475190604, | |
| "learning_rate": 9.642367255058767e-06, | |
| "loss": 0.6454, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.21202361782071927, | |
| "grad_norm": 2.6006832067968717, | |
| "learning_rate": 9.624759327900131e-06, | |
| "loss": 0.6277, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.2147074610842727, | |
| "grad_norm": 2.4933727441933544, | |
| "learning_rate": 9.606745165418554e-06, | |
| "loss": 0.6182, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.21739130434782608, | |
| "grad_norm": 2.4889746389586156, | |
| "learning_rate": 9.58832634996429e-06, | |
| "loss": 0.6281, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.2200751476113795, | |
| "grad_norm": 2.5219239764412986, | |
| "learning_rate": 9.569504499432005e-06, | |
| "loss": 0.6344, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.2227589908749329, | |
| "grad_norm": 2.3745658821898257, | |
| "learning_rate": 9.550281267118659e-06, | |
| "loss": 0.631, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.22544283413848631, | |
| "grad_norm": 2.2346818056908297, | |
| "learning_rate": 9.530658341578276e-06, | |
| "loss": 0.6204, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.2281266774020397, | |
| "grad_norm": 2.4926265077123744, | |
| "learning_rate": 9.510637446473633e-06, | |
| "loss": 0.6529, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.23081052066559313, | |
| "grad_norm": 2.48802821661669, | |
| "learning_rate": 9.490220340424844e-06, | |
| "loss": 0.6174, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.23349436392914655, | |
| "grad_norm": 2.469556269426697, | |
| "learning_rate": 9.469408816854898e-06, | |
| "loss": 0.6043, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.23617820719269994, | |
| "grad_norm": 2.6864451968577336, | |
| "learning_rate": 9.448204703832102e-06, | |
| "loss": 0.6065, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.23886205045625336, | |
| "grad_norm": 2.4420100642120266, | |
| "learning_rate": 9.426609863909537e-06, | |
| "loss": 0.6282, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.24154589371980675, | |
| "grad_norm": 2.510650364856611, | |
| "learning_rate": 9.404626193961427e-06, | |
| "loss": 0.5997, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.24422973698336017, | |
| "grad_norm": 2.3827462346485406, | |
| "learning_rate": 9.382255625016527e-06, | |
| "loss": 0.6008, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.24691358024691357, | |
| "grad_norm": 2.404088618249519, | |
| "learning_rate": 9.359500122088511e-06, | |
| "loss": 0.5927, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.249597423510467, | |
| "grad_norm": 2.3314824711234494, | |
| "learning_rate": 9.336361684003353e-06, | |
| "loss": 0.6096, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.2522812667740204, | |
| "grad_norm": 2.446913723709488, | |
| "learning_rate": 9.312842343223764e-06, | |
| "loss": 0.6106, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.2549651100375738, | |
| "grad_norm": 2.3167110287019703, | |
| "learning_rate": 9.288944165670651e-06, | |
| "loss": 0.599, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.2576489533011272, | |
| "grad_norm": 2.6796711972968565, | |
| "learning_rate": 9.264669250541658e-06, | |
| "loss": 0.6187, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.2603327965646806, | |
| "grad_norm": 2.447285579898827, | |
| "learning_rate": 9.240019730126764e-06, | |
| "loss": 0.6002, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.26301663982823403, | |
| "grad_norm": 2.302643028031218, | |
| "learning_rate": 9.214997769620998e-06, | |
| "loss": 0.5983, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.26570048309178745, | |
| "grad_norm": 2.3880478834994663, | |
| "learning_rate": 9.189605566934235e-06, | |
| "loss": 0.5927, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.2683843263553409, | |
| "grad_norm": 2.2821102026944713, | |
| "learning_rate": 9.163845352498141e-06, | |
| "loss": 0.5896, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.27106816961889424, | |
| "grad_norm": 2.316076514886517, | |
| "learning_rate": 9.137719389070259e-06, | |
| "loss": 0.612, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.27375201288244766, | |
| "grad_norm": 2.3669071775322403, | |
| "learning_rate": 9.111229971535231e-06, | |
| "loss": 0.5921, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.2764358561460011, | |
| "grad_norm": 2.40607674649245, | |
| "learning_rate": 9.084379426703245e-06, | |
| "loss": 0.5938, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.2791196994095545, | |
| "grad_norm": 2.483984679590074, | |
| "learning_rate": 9.057170113105622e-06, | |
| "loss": 0.5789, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.28180354267310787, | |
| "grad_norm": 2.4238872137068133, | |
| "learning_rate": 9.029604420787666e-06, | |
| "loss": 0.6008, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.2844873859366613, | |
| "grad_norm": 2.383320761851924, | |
| "learning_rate": 9.001684771098709e-06, | |
| "loss": 0.5993, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.2871712292002147, | |
| "grad_norm": 2.512312617472668, | |
| "learning_rate": 8.973413616479429e-06, | |
| "loss": 0.5958, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.2898550724637681, | |
| "grad_norm": 2.6864702917091523, | |
| "learning_rate": 8.944793440246435e-06, | |
| "loss": 0.5821, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.29253891572732155, | |
| "grad_norm": 2.489768283560843, | |
| "learning_rate": 8.915826756374118e-06, | |
| "loss": 0.6023, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.2952227589908749, | |
| "grad_norm": 2.380432761185623, | |
| "learning_rate": 8.88651610927384e-06, | |
| "loss": 0.5814, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.29790660225442833, | |
| "grad_norm": 2.462661705077354, | |
| "learning_rate": 8.856864073570429e-06, | |
| "loss": 0.5805, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.30059044551798175, | |
| "grad_norm": 2.486846509153566, | |
| "learning_rate": 8.82687325387603e-06, | |
| "loss": 0.5611, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.3032742887815352, | |
| "grad_norm": 2.541139478220017, | |
| "learning_rate": 8.796546284561307e-06, | |
| "loss": 0.5655, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.3059581320450886, | |
| "grad_norm": 2.379694817443973, | |
| "learning_rate": 8.765885829524059e-06, | |
| "loss": 0.5675, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.30864197530864196, | |
| "grad_norm": 2.350864021403126, | |
| "learning_rate": 8.734894581955208e-06, | |
| "loss": 0.5821, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.3113258185721954, | |
| "grad_norm": 2.379951758536895, | |
| "learning_rate": 8.703575264102245e-06, | |
| "loss": 0.5698, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.3140096618357488, | |
| "grad_norm": 2.3798369127801515, | |
| "learning_rate": 8.6719306270301e-06, | |
| "loss": 0.5631, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.3166935050993022, | |
| "grad_norm": 2.379220740648677, | |
| "learning_rate": 8.639963450379494e-06, | |
| "loss": 0.5751, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.3193773483628556, | |
| "grad_norm": 2.3492048464632784, | |
| "learning_rate": 8.607676542122782e-06, | |
| "loss": 0.5539, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.322061191626409, | |
| "grad_norm": 2.6113830025975173, | |
| "learning_rate": 8.575072738317297e-06, | |
| "loss": 0.5751, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.3247450348899624, | |
| "grad_norm": 2.4877003224906735, | |
| "learning_rate": 8.542154902856232e-06, | |
| "loss": 0.5721, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.32742887815351585, | |
| "grad_norm": 2.3397649413444053, | |
| "learning_rate": 8.508925927217083e-06, | |
| "loss": 0.5621, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.33011272141706927, | |
| "grad_norm": 2.2680164614609133, | |
| "learning_rate": 8.475388730207662e-06, | |
| "loss": 0.5798, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.33279656468062263, | |
| "grad_norm": 2.4317631411749927, | |
| "learning_rate": 8.441546257709708e-06, | |
| "loss": 0.5472, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.33548040794417605, | |
| "grad_norm": 2.5007067593128274, | |
| "learning_rate": 8.407401482420129e-06, | |
| "loss": 0.5534, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.33816425120772947, | |
| "grad_norm": 2.3857521002907447, | |
| "learning_rate": 8.372957403589873e-06, | |
| "loss": 0.5375, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.3408480944712829, | |
| "grad_norm": 2.516555809538038, | |
| "learning_rate": 8.33821704676049e-06, | |
| "loss": 0.5611, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.3435319377348363, | |
| "grad_norm": 2.5676610845279257, | |
| "learning_rate": 8.303183463498357e-06, | |
| "loss": 0.5473, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.3462157809983897, | |
| "grad_norm": 2.302836954637436, | |
| "learning_rate": 8.26785973112664e-06, | |
| "loss": 0.5459, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.3488996242619431, | |
| "grad_norm": 2.3119758449459584, | |
| "learning_rate": 8.232248952454978e-06, | |
| "loss": 0.5442, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.3515834675254965, | |
| "grad_norm": 2.5175857720538817, | |
| "learning_rate": 8.196354255506937e-06, | |
| "loss": 0.5387, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.35426731078904994, | |
| "grad_norm": 2.451900935657762, | |
| "learning_rate": 8.160178793245254e-06, | |
| "loss": 0.5444, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.3569511540526033, | |
| "grad_norm": 2.636025059592631, | |
| "learning_rate": 8.12372574329487e-06, | |
| "loss": 0.5563, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.3596349973161567, | |
| "grad_norm": 2.363196861931475, | |
| "learning_rate": 8.086998307663815e-06, | |
| "loss": 0.5443, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.36231884057971014, | |
| "grad_norm": 2.39932692130308, | |
| "learning_rate": 8.049999712461956e-06, | |
| "loss": 0.5505, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.36500268384326356, | |
| "grad_norm": 2.5554774070794943, | |
| "learning_rate": 8.012733207617602e-06, | |
| "loss": 0.5489, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.367686527106817, | |
| "grad_norm": 2.4170056271369957, | |
| "learning_rate": 7.975202066592038e-06, | |
| "loss": 0.5406, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.37037037037037035, | |
| "grad_norm": 2.494465130607427, | |
| "learning_rate": 7.937409586091996e-06, | |
| "loss": 0.5392, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.37305421363392377, | |
| "grad_norm": 2.4001092959702164, | |
| "learning_rate": 7.899359085780062e-06, | |
| "loss": 0.5434, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.3757380568974772, | |
| "grad_norm": 2.494767087344387, | |
| "learning_rate": 7.86105390798309e-06, | |
| "loss": 0.5442, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.3784219001610306, | |
| "grad_norm": 2.3284854696704356, | |
| "learning_rate": 7.822497417398607e-06, | |
| "loss": 0.5353, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.38110574342458403, | |
| "grad_norm": 2.6279778731831027, | |
| "learning_rate": 7.783693000799263e-06, | |
| "loss": 0.5469, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.3837895866881374, | |
| "grad_norm": 2.4030914588904815, | |
| "learning_rate": 7.744644066735335e-06, | |
| "loss": 0.53, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.3864734299516908, | |
| "grad_norm": 2.568833557957464, | |
| "learning_rate": 7.705354045235335e-06, | |
| "loss": 0.5345, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.38915727321524424, | |
| "grad_norm": 2.7653994763474463, | |
| "learning_rate": 7.665826387504708e-06, | |
| "loss": 0.5393, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.39184111647879766, | |
| "grad_norm": 2.414529557123253, | |
| "learning_rate": 7.626064565622684e-06, | |
| "loss": 0.5116, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.394524959742351, | |
| "grad_norm": 2.4526857440224332, | |
| "learning_rate": 7.586072072237291e-06, | |
| "loss": 0.5119, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.39720880300590444, | |
| "grad_norm": 2.2455030177617026, | |
| "learning_rate": 7.545852420258566e-06, | |
| "loss": 0.5284, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.39989264626945786, | |
| "grad_norm": 2.331410743435388, | |
| "learning_rate": 7.505409142549987e-06, | |
| "loss": 0.5229, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.4025764895330113, | |
| "grad_norm": 2.3706345917751306, | |
| "learning_rate": 7.464745791618138e-06, | |
| "loss": 0.532, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.4052603327965647, | |
| "grad_norm": 2.31993462909309, | |
| "learning_rate": 7.423865939300674e-06, | |
| "loss": 0.5282, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.40794417606011807, | |
| "grad_norm": 2.430242372539381, | |
| "learning_rate": 7.382773176452561e-06, | |
| "loss": 0.5056, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.4106280193236715, | |
| "grad_norm": 2.4104488535149384, | |
| "learning_rate": 7.341471112630667e-06, | |
| "loss": 0.5136, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.4133118625872249, | |
| "grad_norm": 2.390074462196406, | |
| "learning_rate": 7.2999633757766956e-06, | |
| "loss": 0.5247, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.41599570585077833, | |
| "grad_norm": 2.4814551853144797, | |
| "learning_rate": 7.258253611898509e-06, | |
| "loss": 0.5069, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.41867954911433175, | |
| "grad_norm": 2.4009392363914994, | |
| "learning_rate": 7.216345484749876e-06, | |
| "loss": 0.497, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.4213633923778851, | |
| "grad_norm": 2.41177815934589, | |
| "learning_rate": 7.174242675508636e-06, | |
| "loss": 0.5166, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.42404723564143854, | |
| "grad_norm": 2.3766268166919224, | |
| "learning_rate": 7.131948882453361e-06, | |
| "loss": 0.508, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.42673107890499196, | |
| "grad_norm": 2.431582623601011, | |
| "learning_rate": 7.089467820638491e-06, | |
| "loss": 0.5057, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.4294149221685454, | |
| "grad_norm": 2.311511740768648, | |
| "learning_rate": 7.0468032215680116e-06, | |
| "loss": 0.5043, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.43209876543209874, | |
| "grad_norm": 2.5497494220264443, | |
| "learning_rate": 7.003958832867681e-06, | |
| "loss": 0.519, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.43478260869565216, | |
| "grad_norm": 2.264530095456177, | |
| "learning_rate": 6.960938417955841e-06, | |
| "loss": 0.4932, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.4374664519592056, | |
| "grad_norm": 2.2512135660889907, | |
| "learning_rate": 6.917745755712839e-06, | |
| "loss": 0.4996, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.440150295222759, | |
| "grad_norm": 2.513756866525175, | |
| "learning_rate": 6.874384640149101e-06, | |
| "loss": 0.5129, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.4428341384863124, | |
| "grad_norm": 2.3146902188576406, | |
| "learning_rate": 6.830858880071859e-06, | |
| "loss": 0.5012, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.4455179817498658, | |
| "grad_norm": 2.2871230604131125, | |
| "learning_rate": 6.7871722987505975e-06, | |
| "loss": 0.4825, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.4482018250134192, | |
| "grad_norm": 2.368712195841508, | |
| "learning_rate": 6.743328733581211e-06, | |
| "loss": 0.4989, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.45088566827697263, | |
| "grad_norm": 2.3364375835277253, | |
| "learning_rate": 6.6993320357489396e-06, | |
| "loss": 0.488, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.45356951154052605, | |
| "grad_norm": 2.435502626013091, | |
| "learning_rate": 6.6551860698900804e-06, | |
| "loss": 0.5068, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.4562533548040794, | |
| "grad_norm": 2.2559678961633782, | |
| "learning_rate": 6.610894713752513e-06, | |
| "loss": 0.4978, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.45893719806763283, | |
| "grad_norm": 2.421708585514288, | |
| "learning_rate": 6.566461857855096e-06, | |
| "loss": 0.4789, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.46162104133118625, | |
| "grad_norm": 2.3248078025807932, | |
| "learning_rate": 6.52189140514591e-06, | |
| "loss": 0.4657, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.4643048845947397, | |
| "grad_norm": 2.4768863706156368, | |
| "learning_rate": 6.477187270659444e-06, | |
| "loss": 0.4947, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.4669887278582931, | |
| "grad_norm": 2.5478167678970807, | |
| "learning_rate": 6.432353381172686e-06, | |
| "loss": 0.4775, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.46967257112184646, | |
| "grad_norm": 2.339327268365426, | |
| "learning_rate": 6.387393674860205e-06, | |
| "loss": 0.4756, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.4723564143853999, | |
| "grad_norm": 2.314702002618683, | |
| "learning_rate": 6.3423121009482226e-06, | |
| "loss": 0.4919, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.4750402576489533, | |
| "grad_norm": 2.682455259204783, | |
| "learning_rate": 6.297112619367721e-06, | |
| "loss": 0.4783, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.4777241009125067, | |
| "grad_norm": 2.4283892366309754, | |
| "learning_rate": 6.2517992004066e-06, | |
| "loss": 0.4804, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.48040794417606014, | |
| "grad_norm": 2.489416220843048, | |
| "learning_rate": 6.2063758243609275e-06, | |
| "loss": 0.487, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.4830917874396135, | |
| "grad_norm": 2.36035943057164, | |
| "learning_rate": 6.160846481185326e-06, | |
| "loss": 0.4717, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.4857756307031669, | |
| "grad_norm": 2.4213941404437236, | |
| "learning_rate": 6.115215170142481e-06, | |
| "loss": 0.4851, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.48845947396672035, | |
| "grad_norm": 2.3678789802652513, | |
| "learning_rate": 6.069485899451863e-06, | |
| "loss": 0.4696, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.49114331723027377, | |
| "grad_norm": 2.3661062299317503, | |
| "learning_rate": 6.023662685937643e-06, | |
| "loss": 0.4675, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.49382716049382713, | |
| "grad_norm": 2.340045796281873, | |
| "learning_rate": 5.9777495546758556e-06, | |
| "loss": 0.4867, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.49651100375738055, | |
| "grad_norm": 2.2921294482260124, | |
| "learning_rate": 5.9317505386408436e-06, | |
| "loss": 0.4647, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.499194847020934, | |
| "grad_norm": 2.3355811470336163, | |
| "learning_rate": 5.885669678350997e-06, | |
| "loss": 0.4594, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.5018786902844874, | |
| "grad_norm": 2.293840228318117, | |
| "learning_rate": 5.839511021513853e-06, | |
| "loss": 0.4726, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.5045625335480408, | |
| "grad_norm": 2.3615251680146123, | |
| "learning_rate": 5.793278622670527e-06, | |
| "loss": 0.4676, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.5072463768115942, | |
| "grad_norm": 2.456691057919154, | |
| "learning_rate": 5.746976542839583e-06, | |
| "loss": 0.4729, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.5099302200751477, | |
| "grad_norm": 2.492675195450795, | |
| "learning_rate": 5.700608849160307e-06, | |
| "loss": 0.4643, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.512614063338701, | |
| "grad_norm": 2.3999788691818322, | |
| "learning_rate": 5.654179614535457e-06, | |
| "loss": 0.4659, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.5152979066022544, | |
| "grad_norm": 2.36631780419252, | |
| "learning_rate": 5.607692917273499e-06, | |
| "loss": 0.4711, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.5179817498658078, | |
| "grad_norm": 2.290068846637445, | |
| "learning_rate": 5.561152840730371e-06, | |
| "loss": 0.4578, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.5206655931293612, | |
| "grad_norm": 2.304824595871532, | |
| "learning_rate": 5.514563472950812e-06, | |
| "loss": 0.4619, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.5233494363929146, | |
| "grad_norm": 2.4447006864880287, | |
| "learning_rate": 5.46792890630926e-06, | |
| "loss": 0.4565, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.5260332796564681, | |
| "grad_norm": 2.3267145652534875, | |
| "learning_rate": 5.421253237150389e-06, | |
| "loss": 0.4558, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.5287171229200215, | |
| "grad_norm": 2.575226222205548, | |
| "learning_rate": 5.374540565429288e-06, | |
| "loss": 0.4624, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.5314009661835749, | |
| "grad_norm": 2.245840595082055, | |
| "learning_rate": 5.327794994351321e-06, | |
| "loss": 0.4647, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.5340848094471283, | |
| "grad_norm": 2.300998826501448, | |
| "learning_rate": 5.281020630011703e-06, | |
| "loss": 0.4453, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.5367686527106817, | |
| "grad_norm": 2.2487781596141097, | |
| "learning_rate": 5.234221581034834e-06, | |
| "loss": 0.4391, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.5394524959742351, | |
| "grad_norm": 2.2692512235143196, | |
| "learning_rate": 5.187401958213391e-06, | |
| "loss": 0.4399, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.5421363392377885, | |
| "grad_norm": 2.3096587933439876, | |
| "learning_rate": 5.140565874147236e-06, | |
| "loss": 0.4475, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.5448201825013419, | |
| "grad_norm": 2.2597937178702354, | |
| "learning_rate": 5.093717442882185e-06, | |
| "loss": 0.4494, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.5475040257648953, | |
| "grad_norm": 2.47473851041556, | |
| "learning_rate": 5.046860779548613e-06, | |
| "loss": 0.4433, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.5501878690284487, | |
| "grad_norm": 2.3270591862483947, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4428, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.5528717122920022, | |
| "grad_norm": 2.3999930253371535, | |
| "learning_rate": 4.953139220451386e-06, | |
| "loss": 0.4516, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 2.340485387833276, | |
| "learning_rate": 4.906282557117817e-06, | |
| "loss": 0.4339, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.558239398819109, | |
| "grad_norm": 2.304821557482998, | |
| "learning_rate": 4.859434125852764e-06, | |
| "loss": 0.435, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.5609232420826624, | |
| "grad_norm": 2.3444918925482248, | |
| "learning_rate": 4.812598041786611e-06, | |
| "loss": 0.4541, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.5636070853462157, | |
| "grad_norm": 2.4015665488748787, | |
| "learning_rate": 4.765778418965168e-06, | |
| "loss": 0.4468, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.5662909286097692, | |
| "grad_norm": 2.2055102443205574, | |
| "learning_rate": 4.718979369988299e-06, | |
| "loss": 0.4338, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.5689747718733226, | |
| "grad_norm": 2.27457679282764, | |
| "learning_rate": 4.672205005648681e-06, | |
| "loss": 0.4413, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.571658615136876, | |
| "grad_norm": 2.363170032180882, | |
| "learning_rate": 4.625459434570712e-06, | |
| "loss": 0.429, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.5743424584004294, | |
| "grad_norm": 2.4248656899472394, | |
| "learning_rate": 4.578746762849612e-06, | |
| "loss": 0.4439, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.5770263016639828, | |
| "grad_norm": 2.1653108819358904, | |
| "learning_rate": 4.532071093690741e-06, | |
| "loss": 0.4186, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.5797101449275363, | |
| "grad_norm": 2.3722508240459277, | |
| "learning_rate": 4.485436527049189e-06, | |
| "loss": 0.4398, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.5823939881910897, | |
| "grad_norm": 2.241281795727555, | |
| "learning_rate": 4.438847159269631e-06, | |
| "loss": 0.4361, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.5850778314546431, | |
| "grad_norm": 2.331466197551535, | |
| "learning_rate": 4.392307082726503e-06, | |
| "loss": 0.4349, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.5877616747181964, | |
| "grad_norm": 2.367503420958147, | |
| "learning_rate": 4.345820385464543e-06, | |
| "loss": 0.4395, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.5904455179817498, | |
| "grad_norm": 2.152302107871219, | |
| "learning_rate": 4.299391150839694e-06, | |
| "loss": 0.4241, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.5931293612453032, | |
| "grad_norm": 2.295482019593817, | |
| "learning_rate": 4.253023457160418e-06, | |
| "loss": 0.4324, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.5958132045088567, | |
| "grad_norm": 2.1877163583371178, | |
| "learning_rate": 4.2067213773294745e-06, | |
| "loss": 0.4092, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.5984970477724101, | |
| "grad_norm": 2.391292467503754, | |
| "learning_rate": 4.16048897848615e-06, | |
| "loss": 0.4268, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.6011808910359635, | |
| "grad_norm": 2.320260249063, | |
| "learning_rate": 4.114330321649004e-06, | |
| "loss": 0.4298, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.6038647342995169, | |
| "grad_norm": 2.3605082265783817, | |
| "learning_rate": 4.068249461359158e-06, | |
| "loss": 0.4152, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.6065485775630703, | |
| "grad_norm": 2.2129981944822132, | |
| "learning_rate": 4.022250445324145e-06, | |
| "loss": 0.4102, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.6092324208266238, | |
| "grad_norm": 2.240102819743465, | |
| "learning_rate": 3.976337314062358e-06, | |
| "loss": 0.4074, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 0.6119162640901772, | |
| "grad_norm": 2.404806532481161, | |
| "learning_rate": 3.9305141005481385e-06, | |
| "loss": 0.4149, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.6146001073537305, | |
| "grad_norm": 2.22962061461272, | |
| "learning_rate": 3.8847848298575205e-06, | |
| "loss": 0.4167, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 0.6172839506172839, | |
| "grad_norm": 2.4111798540378784, | |
| "learning_rate": 3.839153518814677e-06, | |
| "loss": 0.4055, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.6199677938808373, | |
| "grad_norm": 2.218324658879466, | |
| "learning_rate": 3.7936241756390746e-06, | |
| "loss": 0.3929, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.6226516371443908, | |
| "grad_norm": 2.1673253900912006, | |
| "learning_rate": 3.748200799593402e-06, | |
| "loss": 0.4122, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.6253354804079442, | |
| "grad_norm": 2.2854333924277603, | |
| "learning_rate": 3.7028873806322797e-06, | |
| "loss": 0.3985, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 0.6280193236714976, | |
| "grad_norm": 2.2946578702265414, | |
| "learning_rate": 3.6576878990517783e-06, | |
| "loss": 0.4023, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.630703166935051, | |
| "grad_norm": 2.342820467507063, | |
| "learning_rate": 3.6126063251397968e-06, | |
| "loss": 0.401, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.6333870101986044, | |
| "grad_norm": 2.2381136311788206, | |
| "learning_rate": 3.567646618827315e-06, | |
| "loss": 0.4063, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.6360708534621579, | |
| "grad_norm": 2.2172118668312977, | |
| "learning_rate": 3.522812729340557e-06, | |
| "loss": 0.4145, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 0.6387546967257112, | |
| "grad_norm": 2.2953306356598415, | |
| "learning_rate": 3.4781085948540907e-06, | |
| "loss": 0.4172, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.6414385399892646, | |
| "grad_norm": 2.1880739031851406, | |
| "learning_rate": 3.4335381421449056e-06, | |
| "loss": 0.4117, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 0.644122383252818, | |
| "grad_norm": 2.195458309784443, | |
| "learning_rate": 3.3891052862474884e-06, | |
| "loss": 0.4023, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.6468062265163714, | |
| "grad_norm": 2.321321983061109, | |
| "learning_rate": 3.3448139301099212e-06, | |
| "loss": 0.3889, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 0.6494900697799249, | |
| "grad_norm": 2.3400054868630535, | |
| "learning_rate": 3.3006679642510596e-06, | |
| "loss": 0.4011, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.6521739130434783, | |
| "grad_norm": 2.215041703092911, | |
| "learning_rate": 3.2566712664187907e-06, | |
| "loss": 0.3922, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 0.6548577563070317, | |
| "grad_norm": 2.2784248834066583, | |
| "learning_rate": 3.2128277012494046e-06, | |
| "loss": 0.4019, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.6575415995705851, | |
| "grad_norm": 2.2929065226510965, | |
| "learning_rate": 3.169141119928142e-06, | |
| "loss": 0.401, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.6602254428341385, | |
| "grad_norm": 2.3668328991014507, | |
| "learning_rate": 3.125615359850902e-06, | |
| "loss": 0.401, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.6629092860976918, | |
| "grad_norm": 2.3032852109673936, | |
| "learning_rate": 3.082254244287163e-06, | |
| "loss": 0.4129, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 0.6655931293612453, | |
| "grad_norm": 2.3745775747576974, | |
| "learning_rate": 3.0390615820441605e-06, | |
| "loss": 0.4033, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.6682769726247987, | |
| "grad_norm": 2.112187820926384, | |
| "learning_rate": 2.9960411671323215e-06, | |
| "loss": 0.3918, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 0.6709608158883521, | |
| "grad_norm": 2.352079569297887, | |
| "learning_rate": 2.9531967784319897e-06, | |
| "loss": 0.3913, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.6736446591519055, | |
| "grad_norm": 2.1636433013122134, | |
| "learning_rate": 2.9105321793615106e-06, | |
| "loss": 0.4021, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 0.6763285024154589, | |
| "grad_norm": 2.2820336209542664, | |
| "learning_rate": 2.8680511175466384e-06, | |
| "loss": 0.3969, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.6790123456790124, | |
| "grad_norm": 2.278498833748063, | |
| "learning_rate": 2.8257573244913637e-06, | |
| "loss": 0.3877, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 0.6816961889425658, | |
| "grad_norm": 2.0377851746058693, | |
| "learning_rate": 2.783654515250126e-06, | |
| "loss": 0.3878, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.6843800322061192, | |
| "grad_norm": 2.2235500451646097, | |
| "learning_rate": 2.741746388101493e-06, | |
| "loss": 0.388, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.6870638754696726, | |
| "grad_norm": 2.1172872875998876, | |
| "learning_rate": 2.700036624223308e-06, | |
| "loss": 0.3799, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.6897477187332259, | |
| "grad_norm": 2.1105908414565855, | |
| "learning_rate": 2.6585288873693355e-06, | |
| "loss": 0.3857, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 0.6924315619967794, | |
| "grad_norm": 2.1730901974375887, | |
| "learning_rate": 2.6172268235474396e-06, | |
| "loss": 0.3771, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.6951154052603328, | |
| "grad_norm": 2.276400786258313, | |
| "learning_rate": 2.576134060699328e-06, | |
| "loss": 0.3804, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 0.6977992485238862, | |
| "grad_norm": 2.2639567967043055, | |
| "learning_rate": 2.5352542083818636e-06, | |
| "loss": 0.3683, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.7004830917874396, | |
| "grad_norm": 2.071144906949053, | |
| "learning_rate": 2.494590857450016e-06, | |
| "loss": 0.3929, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 0.703166935050993, | |
| "grad_norm": 2.186986904258155, | |
| "learning_rate": 2.454147579741436e-06, | |
| "loss": 0.3662, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.7058507783145465, | |
| "grad_norm": 2.289270887127177, | |
| "learning_rate": 2.4139279277627113e-06, | |
| "loss": 0.391, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 0.7085346215780999, | |
| "grad_norm": 2.193754293209221, | |
| "learning_rate": 2.373935434377316e-06, | |
| "loss": 0.3849, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.7112184648416533, | |
| "grad_norm": 2.2605664332312263, | |
| "learning_rate": 2.3341736124952924e-06, | |
| "loss": 0.3819, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.7139023081052066, | |
| "grad_norm": 2.1311302604784355, | |
| "learning_rate": 2.294645954764666e-06, | |
| "loss": 0.3782, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.71658615136876, | |
| "grad_norm": 2.1479733454826286, | |
| "learning_rate": 2.2553559332646675e-06, | |
| "loss": 0.376, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 0.7192699946323134, | |
| "grad_norm": 2.4407220469260458, | |
| "learning_rate": 2.216306999200739e-06, | |
| "loss": 0.3804, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.7219538378958669, | |
| "grad_norm": 2.1533154950669346, | |
| "learning_rate": 2.1775025826013953e-06, | |
| "loss": 0.3699, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 0.7246376811594203, | |
| "grad_norm": 2.1200222517367875, | |
| "learning_rate": 2.1389460920169105e-06, | |
| "loss": 0.3758, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.7273215244229737, | |
| "grad_norm": 2.2280252426248546, | |
| "learning_rate": 2.100640914219939e-06, | |
| "loss": 0.3768, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 0.7300053676865271, | |
| "grad_norm": 2.22098019422805, | |
| "learning_rate": 2.062590413908007e-06, | |
| "loss": 0.3817, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.7326892109500805, | |
| "grad_norm": 2.061290288151344, | |
| "learning_rate": 2.0247979334079646e-06, | |
| "loss": 0.3772, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 0.735373054213634, | |
| "grad_norm": 2.1068703177045762, | |
| "learning_rate": 1.9872667923823996e-06, | |
| "loss": 0.3765, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.7380568974771873, | |
| "grad_norm": 2.1721989348036135, | |
| "learning_rate": 1.9500002875380458e-06, | |
| "loss": 0.3693, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "grad_norm": 2.3087123408873724, | |
| "learning_rate": 1.9130016923361843e-06, | |
| "loss": 0.3689, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.7434245840042941, | |
| "grad_norm": 2.249787298183903, | |
| "learning_rate": 1.8762742567051318e-06, | |
| "loss": 0.3697, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 0.7461084272678475, | |
| "grad_norm": 2.0844944176269906, | |
| "learning_rate": 1.8398212067547455e-06, | |
| "loss": 0.3745, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.748792270531401, | |
| "grad_norm": 2.1418534385504313, | |
| "learning_rate": 1.8036457444930643e-06, | |
| "loss": 0.3678, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 0.7514761137949544, | |
| "grad_norm": 2.2265258176825315, | |
| "learning_rate": 1.7677510475450244e-06, | |
| "loss": 0.3712, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.7541599570585078, | |
| "grad_norm": 2.114356169386046, | |
| "learning_rate": 1.7321402688733618e-06, | |
| "loss": 0.3722, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 0.7568438003220612, | |
| "grad_norm": 1.9059264440419736, | |
| "learning_rate": 1.6968165365016443e-06, | |
| "loss": 0.3534, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.7595276435856146, | |
| "grad_norm": 2.0817324970106945, | |
| "learning_rate": 1.661782953239512e-06, | |
| "loss": 0.3553, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 0.7622114868491681, | |
| "grad_norm": 2.1845112300920677, | |
| "learning_rate": 1.6270425964101272e-06, | |
| "loss": 0.3607, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.7648953301127214, | |
| "grad_norm": 2.2388347953996783, | |
| "learning_rate": 1.5925985175798741e-06, | |
| "loss": 0.3631, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.7675791733762748, | |
| "grad_norm": 2.065438224056232, | |
| "learning_rate": 1.5584537422902923e-06, | |
| "loss": 0.3579, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.7702630166398282, | |
| "grad_norm": 2.1879791180316066, | |
| "learning_rate": 1.5246112697923389e-06, | |
| "loss": 0.3623, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 0.7729468599033816, | |
| "grad_norm": 2.12952418145758, | |
| "learning_rate": 1.4910740727829176e-06, | |
| "loss": 0.3636, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.775630703166935, | |
| "grad_norm": 2.105796962684848, | |
| "learning_rate": 1.457845097143769e-06, | |
| "loss": 0.3626, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 0.7783145464304885, | |
| "grad_norm": 2.329028744703023, | |
| "learning_rate": 1.424927261682703e-06, | |
| "loss": 0.3696, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.7809983896940419, | |
| "grad_norm": 2.047922464417174, | |
| "learning_rate": 1.3923234578772177e-06, | |
| "loss": 0.3566, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 0.7836822329575953, | |
| "grad_norm": 2.1754842917957835, | |
| "learning_rate": 1.3600365496205059e-06, | |
| "loss": 0.3735, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.7863660762211487, | |
| "grad_norm": 2.118594900638682, | |
| "learning_rate": 1.3280693729699018e-06, | |
| "loss": 0.3618, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 0.789049919484702, | |
| "grad_norm": 2.110801980026671, | |
| "learning_rate": 1.2964247358977567e-06, | |
| "loss": 0.35, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.7917337627482555, | |
| "grad_norm": 2.0768554801305057, | |
| "learning_rate": 1.265105418044793e-06, | |
| "loss": 0.3564, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.7944176060118089, | |
| "grad_norm": 2.1130567000820912, | |
| "learning_rate": 1.2341141704759418e-06, | |
| "loss": 0.3506, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.7971014492753623, | |
| "grad_norm": 2.1954723852832556, | |
| "learning_rate": 1.2034537154386933e-06, | |
| "loss": 0.3615, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 0.7997852925389157, | |
| "grad_norm": 2.2313097999235163, | |
| "learning_rate": 1.173126746123972e-06, | |
| "loss": 0.3477, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.8024691358024691, | |
| "grad_norm": 2.113285488388264, | |
| "learning_rate": 1.1431359264295717e-06, | |
| "loss": 0.3444, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 0.8051529790660226, | |
| "grad_norm": 2.0060189321333453, | |
| "learning_rate": 1.113483890726162e-06, | |
| "loss": 0.3437, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.807836822329576, | |
| "grad_norm": 2.105485503159171, | |
| "learning_rate": 1.084173243625884e-06, | |
| "loss": 0.3423, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 0.8105206655931294, | |
| "grad_norm": 2.176309187809922, | |
| "learning_rate": 1.0552065597535671e-06, | |
| "loss": 0.3494, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.8132045088566827, | |
| "grad_norm": 2.0086208292784273, | |
| "learning_rate": 1.0265863835205709e-06, | |
| "loss": 0.3444, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 0.8158883521202361, | |
| "grad_norm": 1.9950951432676176, | |
| "learning_rate": 9.983152289012926e-07, | |
| "loss": 0.3446, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.8185721953837896, | |
| "grad_norm": 2.08600856841006, | |
| "learning_rate": 9.70395579212336e-07, | |
| "loss": 0.3612, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.821256038647343, | |
| "grad_norm": 2.143735872309822, | |
| "learning_rate": 9.428298868943791e-07, | |
| "loss": 0.3715, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.8239398819108964, | |
| "grad_norm": 2.10934496917038, | |
| "learning_rate": 9.15620573296756e-07, | |
| "loss": 0.3534, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 0.8266237251744498, | |
| "grad_norm": 2.2376308831295817, | |
| "learning_rate": 8.887700284647699e-07, | |
| "loss": 0.3394, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.8293075684380032, | |
| "grad_norm": 2.0832551370930723, | |
| "learning_rate": 8.622806109297432e-07, | |
| "loss": 0.3401, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 0.8319914117015567, | |
| "grad_norm": 1.9116815001043985, | |
| "learning_rate": 8.361546475018589e-07, | |
| "loss": 0.3459, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.8346752549651101, | |
| "grad_norm": 2.149460928922857, | |
| "learning_rate": 8.103944330657665e-07, | |
| "loss": 0.3488, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 0.8373590982286635, | |
| "grad_norm": 2.2602227225097766, | |
| "learning_rate": 7.850022303790033e-07, | |
| "loss": 0.3498, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.8400429414922168, | |
| "grad_norm": 2.2072319126824174, | |
| "learning_rate": 7.599802698732356e-07, | |
| "loss": 0.3458, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 0.8427267847557702, | |
| "grad_norm": 1.9190250537971258, | |
| "learning_rate": 7.353307494583445e-07, | |
| "loss": 0.3496, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.8454106280193237, | |
| "grad_norm": 1.9837453146123658, | |
| "learning_rate": 7.1105583432935e-07, | |
| "loss": 0.3494, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 0.8480944712828771, | |
| "grad_norm": 2.076563793537066, | |
| "learning_rate": 6.871576567762373e-07, | |
| "loss": 0.3549, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.8507783145464305, | |
| "grad_norm": 2.114401199157526, | |
| "learning_rate": 6.63638315996647e-07, | |
| "loss": 0.3457, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 0.8534621578099839, | |
| "grad_norm": 2.234829127148538, | |
| "learning_rate": 6.404998779114912e-07, | |
| "loss": 0.3482, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.8561460010735373, | |
| "grad_norm": 2.0954925363629715, | |
| "learning_rate": 6.177443749834743e-07, | |
| "loss": 0.3467, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 0.8588298443370908, | |
| "grad_norm": 1.9325306345136635, | |
| "learning_rate": 5.953738060385761e-07, | |
| "loss": 0.3474, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.8615136876006442, | |
| "grad_norm": 1.9905602992978715, | |
| "learning_rate": 5.733901360904648e-07, | |
| "loss": 0.344, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 0.8641975308641975, | |
| "grad_norm": 2.089193748595787, | |
| "learning_rate": 5.517952961678997e-07, | |
| "loss": 0.3494, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.8668813741277509, | |
| "grad_norm": 2.148799764591342, | |
| "learning_rate": 5.305911831451044e-07, | |
| "loss": 0.3486, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "grad_norm": 2.201627113759263, | |
| "learning_rate": 5.097796595751564e-07, | |
| "loss": 0.3426, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.8722490606548577, | |
| "grad_norm": 2.0275804440717007, | |
| "learning_rate": 4.89362553526368e-07, | |
| "loss": 0.3489, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 0.8749329039184112, | |
| "grad_norm": 2.1099431612889, | |
| "learning_rate": 4.6934165842172476e-07, | |
| "loss": 0.3381, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.8776167471819646, | |
| "grad_norm": 2.066634311475606, | |
| "learning_rate": 4.4971873288134237e-07, | |
| "loss": 0.3434, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 0.880300590445518, | |
| "grad_norm": 2.1806395775911596, | |
| "learning_rate": 4.304955005679962e-07, | |
| "loss": 0.3471, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.8829844337090714, | |
| "grad_norm": 2.0157764736789265, | |
| "learning_rate": 4.1167365003571047e-07, | |
| "loss": 0.3413, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 0.8856682769726248, | |
| "grad_norm": 2.0664947846264274, | |
| "learning_rate": 3.9325483458144684e-07, | |
| "loss": 0.347, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.8883521202361782, | |
| "grad_norm": 2.1108092781662484, | |
| "learning_rate": 3.752406720998691e-07, | |
| "loss": 0.3485, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 0.8910359634997316, | |
| "grad_norm": 1.891928460435428, | |
| "learning_rate": 3.5763274494123424e-07, | |
| "loss": 0.3419, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.893719806763285, | |
| "grad_norm": 2.081642191037629, | |
| "learning_rate": 3.404325997723995e-07, | |
| "loss": 0.3449, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 0.8964036500268384, | |
| "grad_norm": 2.1889482427836997, | |
| "learning_rate": 3.2364174744096587e-07, | |
| "loss": 0.3389, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.8990874932903918, | |
| "grad_norm": 2.0522364841531844, | |
| "learning_rate": 3.072616628425601e-07, | |
| "loss": 0.3464, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 0.9017713365539453, | |
| "grad_norm": 2.061106887915519, | |
| "learning_rate": 2.9129378479129e-07, | |
| "loss": 0.3341, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.9044551798174987, | |
| "grad_norm": 2.2381484074888305, | |
| "learning_rate": 2.7573951589335177e-07, | |
| "loss": 0.3428, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 0.9071390230810521, | |
| "grad_norm": 2.218510869316477, | |
| "learning_rate": 2.60600222423833e-07, | |
| "loss": 0.3436, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.9098228663446055, | |
| "grad_norm": 1.9815557147120695, | |
| "learning_rate": 2.45877234206694e-07, | |
| "loss": 0.3319, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 0.9125067096081588, | |
| "grad_norm": 2.1772783247451044, | |
| "learning_rate": 2.3157184449796365e-07, | |
| "loss": 0.3416, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.9151905528717122, | |
| "grad_norm": 1.981991641534087, | |
| "learning_rate": 2.1768530987213267e-07, | |
| "loss": 0.3388, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 0.9178743961352657, | |
| "grad_norm": 1.9438279915733367, | |
| "learning_rate": 2.0421885011178532e-07, | |
| "loss": 0.3347, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.9205582393988191, | |
| "grad_norm": 2.1255065876738, | |
| "learning_rate": 1.911736481004489e-07, | |
| "loss": 0.3394, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 0.9232420826623725, | |
| "grad_norm": 2.119417709846858, | |
| "learning_rate": 1.7855084971869385e-07, | |
| "loss": 0.3424, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.9259259259259259, | |
| "grad_norm": 1.9161088716481145, | |
| "learning_rate": 1.663515637434765e-07, | |
| "loss": 0.3437, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 0.9286097691894794, | |
| "grad_norm": 2.090108009100012, | |
| "learning_rate": 1.5457686175074816e-07, | |
| "loss": 0.3353, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.9312936124530328, | |
| "grad_norm": 2.1647653121366615, | |
| "learning_rate": 1.432277780213298e-07, | |
| "loss": 0.3363, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 0.9339774557165862, | |
| "grad_norm": 1.896589465841101, | |
| "learning_rate": 1.3230530945005704e-07, | |
| "loss": 0.3323, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.9366612989801396, | |
| "grad_norm": 2.024982025615293, | |
| "learning_rate": 1.2181041545821727e-07, | |
| "loss": 0.345, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 0.9393451422436929, | |
| "grad_norm": 1.9554765281477002, | |
| "learning_rate": 1.1174401790927325e-07, | |
| "loss": 0.3374, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.9420289855072463, | |
| "grad_norm": 2.134402406852014, | |
| "learning_rate": 1.0210700102788796e-07, | |
| "loss": 0.3437, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 0.9447128287707998, | |
| "grad_norm": 1.9646942700384806, | |
| "learning_rate": 9.290021132225546e-08, | |
| "loss": 0.3378, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.9473966720343532, | |
| "grad_norm": 1.9964844201588634, | |
| "learning_rate": 8.412445750974385e-08, | |
| "loss": 0.3301, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 0.9500805152979066, | |
| "grad_norm": 2.064203950697388, | |
| "learning_rate": 7.578051044585644e-08, | |
| "loss": 0.3568, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.95276435856146, | |
| "grad_norm": 1.9613680272123186, | |
| "learning_rate": 6.786910305652373e-08, | |
| "loss": 0.3343, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 0.9554482018250134, | |
| "grad_norm": 2.1016694828101232, | |
| "learning_rate": 6.0390930273721e-08, | |
| "loss": 0.3503, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.9581320450885669, | |
| "grad_norm": 2.1012546083202785, | |
| "learning_rate": 5.334664897442721e-08, | |
| "loss": 0.3361, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 0.9608158883521203, | |
| "grad_norm": 1.9602399940420405, | |
| "learning_rate": 4.673687792292436e-08, | |
| "loss": 0.3353, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.9634997316156736, | |
| "grad_norm": 1.9055816709169238, | |
| "learning_rate": 4.0562197716448316e-08, | |
| "loss": 0.3398, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 0.966183574879227, | |
| "grad_norm": 1.9729031450192134, | |
| "learning_rate": 3.4823150734186184e-08, | |
| "loss": 0.3398, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.9688674181427804, | |
| "grad_norm": 2.0388674602910126, | |
| "learning_rate": 2.9520241089636692e-08, | |
| "loss": 0.3402, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 0.9715512614063339, | |
| "grad_norm": 2.0386145711059824, | |
| "learning_rate": 2.4653934586328365e-08, | |
| "loss": 0.3321, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.9742351046698873, | |
| "grad_norm": 1.9910102377492196, | |
| "learning_rate": 2.022465867690282e-08, | |
| "loss": 0.3321, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 0.9769189479334407, | |
| "grad_norm": 1.910144000929515, | |
| "learning_rate": 1.6232802425568683e-08, | |
| "loss": 0.3398, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.9796027911969941, | |
| "grad_norm": 1.8586744864697868, | |
| "learning_rate": 1.2678716473927821e-08, | |
| "loss": 0.3244, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 0.9822866344605475, | |
| "grad_norm": 2.0221632303430446, | |
| "learning_rate": 9.562713010171088e-09, | |
| "loss": 0.342, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.984970477724101, | |
| "grad_norm": 2.2840920931201354, | |
| "learning_rate": 6.885065741661367e-09, | |
| "loss": 0.3386, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 0.9876543209876543, | |
| "grad_norm": 2.038298019498587, | |
| "learning_rate": 4.646009870885593e-09, | |
| "loss": 0.3417, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.9903381642512077, | |
| "grad_norm": 2.0920848641927288, | |
| "learning_rate": 2.845742074801261e-09, | |
| "loss": 0.334, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 0.9930220075147611, | |
| "grad_norm": 1.8993207355789097, | |
| "learning_rate": 1.484420487553595e-09, | |
| "loss": 0.3332, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.9957058507783145, | |
| "grad_norm": 2.041055414162317, | |
| "learning_rate": 5.621646865899832e-10, | |
| "loss": 0.337, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 0.998389694041868, | |
| "grad_norm": 2.005150478538608, | |
| "learning_rate": 7.905568215504921e-11, | |
| "loss": 0.3552, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_runtime": 3.3614, | |
| "eval_samples_per_second": 2.975, | |
| "eval_steps_per_second": 0.892, | |
| "step": 1863 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1863, | |
| "total_flos": 195037149265920.0, | |
| "train_loss": 0.502085976490588, | |
| "train_runtime": 16545.7004, | |
| "train_samples_per_second": 1.801, | |
| "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 | |
| } | |