ambrim's picture
Upload 10 files
adadbd5 verified
{
"best_metric": 0.939085176178107,
"best_model_checkpoint": "w-attempt11_150M_200k_weighted_accuracy_3/checkpoint-68620",
"epoch": 19.0,
"eval_steps": 500,
"global_step": 130378,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.07,
"learning_rate": 9.963567473039931e-05,
"loss": -45291724.8,
"step": 500
},
{
"epoch": 0.15,
"learning_rate": 9.92713494607986e-05,
"loss": -45805633.536,
"step": 1000
},
{
"epoch": 0.22,
"learning_rate": 9.890702419119791e-05,
"loss": -45895168.0,
"step": 1500
},
{
"epoch": 0.29,
"learning_rate": 9.854269892159721e-05,
"loss": -46820265.984,
"step": 2000
},
{
"epoch": 0.36,
"learning_rate": 9.81783736519965e-05,
"loss": -45716246.528,
"step": 2500
},
{
"epoch": 0.44,
"learning_rate": 9.781404838239581e-05,
"loss": -47373447.168,
"step": 3000
},
{
"epoch": 0.51,
"learning_rate": 9.744972311279511e-05,
"loss": -46621626.368,
"step": 3500
},
{
"epoch": 0.58,
"learning_rate": 9.70853978431944e-05,
"loss": -46436790.272,
"step": 4000
},
{
"epoch": 0.66,
"learning_rate": 9.672107257359371e-05,
"loss": -48448180.224,
"step": 4500
},
{
"epoch": 0.73,
"learning_rate": 9.635674730399302e-05,
"loss": -49018150.912,
"step": 5000
},
{
"epoch": 0.8,
"learning_rate": 9.599242203439231e-05,
"loss": -47903965.184,
"step": 5500
},
{
"epoch": 0.87,
"learning_rate": 9.562809676479161e-05,
"loss": -48040722.432,
"step": 6000
},
{
"epoch": 0.95,
"learning_rate": 9.52637714951909e-05,
"loss": -48426446.848,
"step": 6500
},
{
"epoch": 1.0,
"eval_accuracy": 0.9390846650576595,
"eval_loss": -48866836.0,
"eval_runtime": 140.4892,
"eval_samples_per_second": 32.565,
"eval_steps_per_second": 5.431,
"step": 6862
},
{
"epoch": 1.02,
"learning_rate": 9.489944622559021e-05,
"loss": -48707878.912,
"step": 7000
},
{
"epoch": 1.09,
"learning_rate": 9.45351209559895e-05,
"loss": -48791568.384,
"step": 7500
},
{
"epoch": 1.17,
"learning_rate": 9.417079568638881e-05,
"loss": -49470013.44,
"step": 8000
},
{
"epoch": 1.24,
"learning_rate": 9.380647041678811e-05,
"loss": -49581268.992,
"step": 8500
},
{
"epoch": 1.31,
"learning_rate": 9.34421451471874e-05,
"loss": -49812631.552,
"step": 9000
},
{
"epoch": 1.38,
"learning_rate": 9.307781987758671e-05,
"loss": -51050233.856,
"step": 9500
},
{
"epoch": 1.46,
"learning_rate": 9.271349460798602e-05,
"loss": -49380384.768,
"step": 10000
},
{
"epoch": 1.53,
"learning_rate": 9.234916933838531e-05,
"loss": -50849955.84,
"step": 10500
},
{
"epoch": 1.6,
"learning_rate": 9.198484406878461e-05,
"loss": -50895675.392,
"step": 11000
},
{
"epoch": 1.68,
"learning_rate": 9.162051879918392e-05,
"loss": -51525443.584,
"step": 11500
},
{
"epoch": 1.75,
"learning_rate": 9.125619352958321e-05,
"loss": -51623043.072,
"step": 12000
},
{
"epoch": 1.82,
"learning_rate": 9.089186825998251e-05,
"loss": -51431817.216,
"step": 12500
},
{
"epoch": 1.89,
"learning_rate": 9.052754299038182e-05,
"loss": -52294926.336,
"step": 13000
},
{
"epoch": 1.97,
"learning_rate": 9.016321772078111e-05,
"loss": -52310294.528,
"step": 13500
},
{
"epoch": 2.0,
"eval_accuracy": 0.9390846650576595,
"eval_loss": -52481936.0,
"eval_runtime": 140.6751,
"eval_samples_per_second": 32.522,
"eval_steps_per_second": 5.424,
"step": 13724
},
{
"epoch": 2.04,
"learning_rate": 8.979889245118042e-05,
"loss": -52849963.008,
"step": 14000
},
{
"epoch": 2.11,
"learning_rate": 8.943456718157972e-05,
"loss": -51996053.504,
"step": 14500
},
{
"epoch": 2.19,
"learning_rate": 8.907024191197903e-05,
"loss": -52932702.208,
"step": 15000
},
{
"epoch": 2.26,
"learning_rate": 8.870591664237832e-05,
"loss": -53120417.792,
"step": 15500
},
{
"epoch": 2.33,
"learning_rate": 8.834159137277762e-05,
"loss": -53047812.096,
"step": 16000
},
{
"epoch": 2.4,
"learning_rate": 8.797726610317693e-05,
"loss": -53340852.224,
"step": 16500
},
{
"epoch": 2.48,
"learning_rate": 8.761294083357622e-05,
"loss": -54068551.68,
"step": 17000
},
{
"epoch": 2.55,
"learning_rate": 8.724861556397553e-05,
"loss": -54157238.272,
"step": 17500
},
{
"epoch": 2.62,
"learning_rate": 8.688429029437483e-05,
"loss": -55157616.64,
"step": 18000
},
{
"epoch": 2.7,
"learning_rate": 8.651996502477412e-05,
"loss": -54994378.752,
"step": 18500
},
{
"epoch": 2.77,
"learning_rate": 8.615563975517343e-05,
"loss": -55051157.504,
"step": 19000
},
{
"epoch": 2.84,
"learning_rate": 8.579131448557272e-05,
"loss": -55962931.2,
"step": 19500
},
{
"epoch": 2.91,
"learning_rate": 8.542698921597203e-05,
"loss": -55627776.0,
"step": 20000
},
{
"epoch": 2.99,
"learning_rate": 8.506266394637132e-05,
"loss": -56315912.192,
"step": 20500
},
{
"epoch": 3.0,
"eval_accuracy": 0.925901846473729,
"eval_loss": -56022420.0,
"eval_runtime": 140.5361,
"eval_samples_per_second": 32.554,
"eval_steps_per_second": 5.429,
"step": 20586
},
{
"epoch": 3.06,
"learning_rate": 8.469833867677062e-05,
"loss": -54493945.856,
"step": 21000
},
{
"epoch": 3.13,
"learning_rate": 8.433401340716991e-05,
"loss": -56493727.744,
"step": 21500
},
{
"epoch": 3.21,
"learning_rate": 8.396968813756922e-05,
"loss": -55263518.72,
"step": 22000
},
{
"epoch": 3.28,
"learning_rate": 8.360536286796853e-05,
"loss": -57487265.792,
"step": 22500
},
{
"epoch": 3.35,
"learning_rate": 8.324103759836782e-05,
"loss": -56985743.36,
"step": 23000
},
{
"epoch": 3.42,
"learning_rate": 8.287671232876712e-05,
"loss": -58815537.152,
"step": 23500
},
{
"epoch": 3.5,
"learning_rate": 8.251238705916643e-05,
"loss": -57407537.152,
"step": 24000
},
{
"epoch": 3.57,
"learning_rate": 8.214806178956573e-05,
"loss": -57564954.624,
"step": 24500
},
{
"epoch": 3.64,
"learning_rate": 8.178373651996502e-05,
"loss": -58936123.392,
"step": 25000
},
{
"epoch": 3.72,
"learning_rate": 8.141941125036433e-05,
"loss": -57919283.2,
"step": 25500
},
{
"epoch": 3.79,
"learning_rate": 8.105508598076364e-05,
"loss": -59147018.24,
"step": 26000
},
{
"epoch": 3.86,
"learning_rate": 8.069076071116293e-05,
"loss": -60339425.28,
"step": 26500
},
{
"epoch": 3.93,
"learning_rate": 8.032643544156223e-05,
"loss": -59097784.32,
"step": 27000
},
{
"epoch": 4.0,
"eval_accuracy": 0.2777975411017508,
"eval_loss": -59458916.0,
"eval_runtime": 140.4799,
"eval_samples_per_second": 32.567,
"eval_steps_per_second": 5.431,
"step": 27448
},
{
"epoch": 4.01,
"learning_rate": 7.996211017196154e-05,
"loss": -59645280.256,
"step": 27500
},
{
"epoch": 4.08,
"learning_rate": 7.959778490236083e-05,
"loss": -59958251.52,
"step": 28000
},
{
"epoch": 4.15,
"learning_rate": 7.923345963276013e-05,
"loss": -59766607.872,
"step": 28500
},
{
"epoch": 4.23,
"learning_rate": 7.886913436315944e-05,
"loss": -59114577.92,
"step": 29000
},
{
"epoch": 4.3,
"learning_rate": 7.850480909355873e-05,
"loss": -60154269.696,
"step": 29500
},
{
"epoch": 4.37,
"learning_rate": 7.814048382395804e-05,
"loss": -60939288.576,
"step": 30000
},
{
"epoch": 4.44,
"learning_rate": 7.777615855435734e-05,
"loss": -61184118.784,
"step": 30500
},
{
"epoch": 4.52,
"learning_rate": 7.741183328475663e-05,
"loss": -60580569.088,
"step": 31000
},
{
"epoch": 4.59,
"learning_rate": 7.704750801515594e-05,
"loss": -61123616.768,
"step": 31500
},
{
"epoch": 4.66,
"learning_rate": 7.668318274555523e-05,
"loss": -61707681.792,
"step": 32000
},
{
"epoch": 4.74,
"learning_rate": 7.631885747595454e-05,
"loss": -62141902.848,
"step": 32500
},
{
"epoch": 4.81,
"learning_rate": 7.595453220635383e-05,
"loss": -62257491.968,
"step": 33000
},
{
"epoch": 4.88,
"learning_rate": 7.559020693675313e-05,
"loss": -61910138.88,
"step": 33500
},
{
"epoch": 4.95,
"learning_rate": 7.522588166715244e-05,
"loss": -62387458.048,
"step": 34000
},
{
"epoch": 5.0,
"eval_accuracy": 0.9374904803816638,
"eval_loss": -62782156.0,
"eval_runtime": 140.6016,
"eval_samples_per_second": 32.539,
"eval_steps_per_second": 5.427,
"step": 34310
},
{
"epoch": 5.03,
"learning_rate": 7.486155639755173e-05,
"loss": -62778101.76,
"step": 34500
},
{
"epoch": 5.1,
"learning_rate": 7.449723112795104e-05,
"loss": -61654474.752,
"step": 35000
},
{
"epoch": 5.17,
"learning_rate": 7.413290585835034e-05,
"loss": -63444623.36,
"step": 35500
},
{
"epoch": 5.25,
"learning_rate": 7.376858058874963e-05,
"loss": -63924170.752,
"step": 36000
},
{
"epoch": 5.32,
"learning_rate": 7.340425531914894e-05,
"loss": -64446738.432,
"step": 36500
},
{
"epoch": 5.39,
"learning_rate": 7.303993004954824e-05,
"loss": -63802613.76,
"step": 37000
},
{
"epoch": 5.46,
"learning_rate": 7.267560477994753e-05,
"loss": -64431120.384,
"step": 37500
},
{
"epoch": 5.54,
"learning_rate": 7.231127951034684e-05,
"loss": -64963448.832,
"step": 38000
},
{
"epoch": 5.61,
"learning_rate": 7.194695424074615e-05,
"loss": -65131409.408,
"step": 38500
},
{
"epoch": 5.68,
"learning_rate": 7.158262897114544e-05,
"loss": -65691656.192,
"step": 39000
},
{
"epoch": 5.76,
"learning_rate": 7.121830370154474e-05,
"loss": -64785408.0,
"step": 39500
},
{
"epoch": 5.83,
"learning_rate": 7.085397843194405e-05,
"loss": -65206128.64,
"step": 40000
},
{
"epoch": 5.9,
"learning_rate": 7.048965316234334e-05,
"loss": -66662268.928,
"step": 40500
},
{
"epoch": 5.97,
"learning_rate": 7.012532789274264e-05,
"loss": -66883559.424,
"step": 41000
},
{
"epoch": 6.0,
"eval_accuracy": 0.0615205015522728,
"eval_loss": -65968124.0,
"eval_runtime": 140.5904,
"eval_samples_per_second": 32.541,
"eval_steps_per_second": 5.427,
"step": 41172
},
{
"epoch": 6.05,
"learning_rate": 6.976100262314195e-05,
"loss": -66262577.152,
"step": 41500
},
{
"epoch": 6.12,
"learning_rate": 6.939667735354124e-05,
"loss": -66138095.616,
"step": 42000
},
{
"epoch": 6.19,
"learning_rate": 6.903235208394055e-05,
"loss": -66000941.056,
"step": 42500
},
{
"epoch": 6.27,
"learning_rate": 6.866802681433985e-05,
"loss": -66043711.488,
"step": 43000
},
{
"epoch": 6.34,
"learning_rate": 6.830370154473916e-05,
"loss": -67294355.456,
"step": 43500
},
{
"epoch": 6.41,
"learning_rate": 6.793937627513845e-05,
"loss": -66696265.728,
"step": 44000
},
{
"epoch": 6.48,
"learning_rate": 6.757505100553775e-05,
"loss": -67892416.512,
"step": 44500
},
{
"epoch": 6.56,
"learning_rate": 6.721072573593705e-05,
"loss": -68610072.576,
"step": 45000
},
{
"epoch": 6.63,
"learning_rate": 6.684640046633635e-05,
"loss": -68491333.632,
"step": 45500
},
{
"epoch": 6.7,
"learning_rate": 6.648207519673564e-05,
"loss": -67069939.712,
"step": 46000
},
{
"epoch": 6.78,
"learning_rate": 6.611774992713495e-05,
"loss": -68588945.408,
"step": 46500
},
{
"epoch": 6.85,
"learning_rate": 6.575342465753424e-05,
"loss": -68691898.368,
"step": 47000
},
{
"epoch": 6.92,
"learning_rate": 6.538909938793355e-05,
"loss": -68925136.896,
"step": 47500
},
{
"epoch": 7.0,
"learning_rate": 6.502477411833285e-05,
"loss": -68470861.824,
"step": 48000
},
{
"epoch": 7.0,
"eval_accuracy": 0.09869071386148431,
"eval_loss": -69001016.0,
"eval_runtime": 140.4928,
"eval_samples_per_second": 32.564,
"eval_steps_per_second": 5.431,
"step": 48034
},
{
"epoch": 7.07,
"learning_rate": 6.466044884873214e-05,
"loss": -69397372.928,
"step": 48500
},
{
"epoch": 7.14,
"learning_rate": 6.429612357913145e-05,
"loss": -68643381.248,
"step": 49000
},
{
"epoch": 7.21,
"learning_rate": 6.393179830953075e-05,
"loss": -69129224.192,
"step": 49500
},
{
"epoch": 7.29,
"learning_rate": 6.356747303993004e-05,
"loss": -68606349.312,
"step": 50000
},
{
"epoch": 7.36,
"learning_rate": 6.320314777032935e-05,
"loss": -69869944.832,
"step": 50500
},
{
"epoch": 7.43,
"learning_rate": 6.283882250072866e-05,
"loss": -69662130.176,
"step": 51000
},
{
"epoch": 7.51,
"learning_rate": 6.247449723112795e-05,
"loss": -70071443.456,
"step": 51500
},
{
"epoch": 7.58,
"learning_rate": 6.211017196152725e-05,
"loss": -71158382.592,
"step": 52000
},
{
"epoch": 7.65,
"learning_rate": 6.174584669192656e-05,
"loss": -71736705.024,
"step": 52500
},
{
"epoch": 7.72,
"learning_rate": 6.138152142232586e-05,
"loss": -71263109.12,
"step": 53000
},
{
"epoch": 7.8,
"learning_rate": 6.1017196152725155e-05,
"loss": -71382908.928,
"step": 53500
},
{
"epoch": 7.87,
"learning_rate": 6.065287088312446e-05,
"loss": -72212193.28,
"step": 54000
},
{
"epoch": 7.94,
"learning_rate": 6.028854561352376e-05,
"loss": -71681441.792,
"step": 54500
},
{
"epoch": 8.0,
"eval_accuracy": 0.9390846650576595,
"eval_loss": -71871320.0,
"eval_runtime": 140.58,
"eval_samples_per_second": 32.544,
"eval_steps_per_second": 5.428,
"step": 54896
},
{
"epoch": 8.02,
"learning_rate": 5.992422034392306e-05,
"loss": -72663113.728,
"step": 55000
},
{
"epoch": 8.09,
"learning_rate": 5.9559895074322355e-05,
"loss": -71982637.056,
"step": 55500
},
{
"epoch": 8.16,
"learning_rate": 5.919556980472166e-05,
"loss": -72483135.488,
"step": 56000
},
{
"epoch": 8.23,
"learning_rate": 5.883124453512095e-05,
"loss": -72597225.472,
"step": 56500
},
{
"epoch": 8.31,
"learning_rate": 5.846691926552026e-05,
"loss": -74057515.008,
"step": 57000
},
{
"epoch": 8.38,
"learning_rate": 5.810259399591956e-05,
"loss": -72980668.416,
"step": 57500
},
{
"epoch": 8.45,
"learning_rate": 5.7738268726318855e-05,
"loss": -72482070.528,
"step": 58000
},
{
"epoch": 8.53,
"learning_rate": 5.737394345671816e-05,
"loss": -73420226.56,
"step": 58500
},
{
"epoch": 8.6,
"learning_rate": 5.7009618187117465e-05,
"loss": -73117335.552,
"step": 59000
},
{
"epoch": 8.67,
"learning_rate": 5.664529291751676e-05,
"loss": -75365900.288,
"step": 59500
},
{
"epoch": 8.74,
"learning_rate": 5.628096764791606e-05,
"loss": -73344704.512,
"step": 60000
},
{
"epoch": 8.82,
"learning_rate": 5.591664237831537e-05,
"loss": -74256801.792,
"step": 60500
},
{
"epoch": 8.89,
"learning_rate": 5.555231710871466e-05,
"loss": -74018381.824,
"step": 61000
},
{
"epoch": 8.96,
"learning_rate": 5.5187991839113965e-05,
"loss": -74225754.112,
"step": 61500
},
{
"epoch": 9.0,
"eval_accuracy": 0.8926442611907266,
"eval_loss": -74553696.0,
"eval_runtime": 140.5145,
"eval_samples_per_second": 32.559,
"eval_steps_per_second": 5.43,
"step": 61758
},
{
"epoch": 9.04,
"learning_rate": 5.482366656951326e-05,
"loss": -74975141.888,
"step": 62000
},
{
"epoch": 9.11,
"learning_rate": 5.445934129991257e-05,
"loss": -75021574.144,
"step": 62500
},
{
"epoch": 9.18,
"learning_rate": 5.409501603031186e-05,
"loss": -74001276.928,
"step": 63000
},
{
"epoch": 9.25,
"learning_rate": 5.3730690760711165e-05,
"loss": -75256430.592,
"step": 63500
},
{
"epoch": 9.33,
"learning_rate": 5.336636549111047e-05,
"loss": -75693981.696,
"step": 64000
},
{
"epoch": 9.4,
"learning_rate": 5.300204022150976e-05,
"loss": -75368284.16,
"step": 64500
},
{
"epoch": 9.47,
"learning_rate": 5.263771495190907e-05,
"loss": -76256927.744,
"step": 65000
},
{
"epoch": 9.55,
"learning_rate": 5.227338968230837e-05,
"loss": -75650670.592,
"step": 65500
},
{
"epoch": 9.62,
"learning_rate": 5.1909064412707665e-05,
"loss": -75926216.704,
"step": 66000
},
{
"epoch": 9.69,
"learning_rate": 5.154473914310697e-05,
"loss": -74388004.864,
"step": 66500
},
{
"epoch": 9.76,
"learning_rate": 5.1180413873506275e-05,
"loss": -75555790.848,
"step": 67000
},
{
"epoch": 9.84,
"learning_rate": 5.081608860390557e-05,
"loss": -76560392.192,
"step": 67500
},
{
"epoch": 9.91,
"learning_rate": 5.045176333430487e-05,
"loss": -76913991.68,
"step": 68000
},
{
"epoch": 9.98,
"learning_rate": 5.008743806470417e-05,
"loss": -77898022.912,
"step": 68500
},
{
"epoch": 10.0,
"eval_accuracy": 0.939085176178107,
"eval_loss": -77044920.0,
"eval_runtime": 140.4212,
"eval_samples_per_second": 32.581,
"eval_steps_per_second": 5.434,
"step": 68620
},
{
"epoch": 10.06,
"learning_rate": 4.972311279510347e-05,
"loss": -77507895.296,
"step": 69000
},
{
"epoch": 10.13,
"learning_rate": 4.935878752550277e-05,
"loss": -77761937.408,
"step": 69500
},
{
"epoch": 10.2,
"learning_rate": 4.899446225590207e-05,
"loss": -76288212.992,
"step": 70000
},
{
"epoch": 10.27,
"learning_rate": 4.863013698630137e-05,
"loss": -77314039.808,
"step": 70500
},
{
"epoch": 10.35,
"learning_rate": 4.826581171670067e-05,
"loss": -78605926.4,
"step": 71000
},
{
"epoch": 10.42,
"learning_rate": 4.7901486447099975e-05,
"loss": -78309654.528,
"step": 71500
},
{
"epoch": 10.49,
"learning_rate": 4.7537161177499274e-05,
"loss": -77582622.72,
"step": 72000
},
{
"epoch": 10.57,
"learning_rate": 4.717283590789857e-05,
"loss": -78851588.096,
"step": 72500
},
{
"epoch": 10.64,
"learning_rate": 4.680851063829788e-05,
"loss": -78221975.552,
"step": 73000
},
{
"epoch": 10.71,
"learning_rate": 4.6444185368697176e-05,
"loss": -78484627.456,
"step": 73500
},
{
"epoch": 10.78,
"learning_rate": 4.6079860099096475e-05,
"loss": -79210758.144,
"step": 74000
},
{
"epoch": 10.86,
"learning_rate": 4.571553482949577e-05,
"loss": -79154405.376,
"step": 74500
},
{
"epoch": 10.93,
"learning_rate": 4.535120955989507e-05,
"loss": -79427207.168,
"step": 75000
},
{
"epoch": 11.0,
"eval_accuracy": 0.9390846650576595,
"eval_loss": -79331408.0,
"eval_runtime": 140.8493,
"eval_samples_per_second": 32.482,
"eval_steps_per_second": 5.417,
"step": 75482
},
{
"epoch": 11.0,
"learning_rate": 4.498688429029438e-05,
"loss": -77644054.528,
"step": 75500
},
{
"epoch": 11.08,
"learning_rate": 4.4622559020693675e-05,
"loss": -80382214.144,
"step": 76000
},
{
"epoch": 11.15,
"learning_rate": 4.4258233751092974e-05,
"loss": -79687761.92,
"step": 76500
},
{
"epoch": 11.22,
"learning_rate": 4.389390848149228e-05,
"loss": -80717864.96,
"step": 77000
},
{
"epoch": 11.29,
"learning_rate": 4.352958321189158e-05,
"loss": -78811938.816,
"step": 77500
},
{
"epoch": 11.37,
"learning_rate": 4.316525794229088e-05,
"loss": -81066213.376,
"step": 78000
},
{
"epoch": 11.44,
"learning_rate": 4.280093267269018e-05,
"loss": -80793083.904,
"step": 78500
},
{
"epoch": 11.51,
"learning_rate": 4.243660740308948e-05,
"loss": -81717043.2,
"step": 79000
},
{
"epoch": 11.59,
"learning_rate": 4.2072282133488785e-05,
"loss": -80945037.312,
"step": 79500
},
{
"epoch": 11.66,
"learning_rate": 4.1707956863888084e-05,
"loss": -80173359.104,
"step": 80000
},
{
"epoch": 11.73,
"learning_rate": 4.134363159428738e-05,
"loss": -81400258.56,
"step": 80500
},
{
"epoch": 11.8,
"learning_rate": 4.097930632468668e-05,
"loss": -81076838.4,
"step": 81000
},
{
"epoch": 11.88,
"learning_rate": 4.061498105508598e-05,
"loss": -80894648.32,
"step": 81500
},
{
"epoch": 11.95,
"learning_rate": 4.025065578548528e-05,
"loss": -81671749.632,
"step": 82000
},
{
"epoch": 12.0,
"eval_accuracy": 0.9085907080347112,
"eval_loss": -81395544.0,
"eval_runtime": 140.5708,
"eval_samples_per_second": 32.546,
"eval_steps_per_second": 5.428,
"step": 82344
},
{
"epoch": 12.02,
"learning_rate": 3.988633051588458e-05,
"loss": -81880604.672,
"step": 82500
},
{
"epoch": 12.1,
"learning_rate": 3.952200524628388e-05,
"loss": -81586765.824,
"step": 83000
},
{
"epoch": 12.17,
"learning_rate": 3.915767997668319e-05,
"loss": -82552373.248,
"step": 83500
},
{
"epoch": 12.24,
"learning_rate": 3.8793354707082485e-05,
"loss": -81833467.904,
"step": 84000
},
{
"epoch": 12.31,
"learning_rate": 3.8429029437481784e-05,
"loss": -81576132.608,
"step": 84500
},
{
"epoch": 12.39,
"learning_rate": 3.806470416788109e-05,
"loss": -81537269.76,
"step": 85000
},
{
"epoch": 12.46,
"learning_rate": 3.770037889828039e-05,
"loss": -82501386.24,
"step": 85500
},
{
"epoch": 12.53,
"learning_rate": 3.7336053628679686e-05,
"loss": -82590752.768,
"step": 86000
},
{
"epoch": 12.61,
"learning_rate": 3.697172835907899e-05,
"loss": -82240012.288,
"step": 86500
},
{
"epoch": 12.68,
"learning_rate": 3.660740308947829e-05,
"loss": -82220097.536,
"step": 87000
},
{
"epoch": 12.75,
"learning_rate": 3.624307781987759e-05,
"loss": -84001292.288,
"step": 87500
},
{
"epoch": 12.82,
"learning_rate": 3.587875255027689e-05,
"loss": -82441928.704,
"step": 88000
},
{
"epoch": 12.9,
"learning_rate": 3.5514427280676185e-05,
"loss": -81264050.176,
"step": 88500
},
{
"epoch": 12.97,
"learning_rate": 3.515010201107549e-05,
"loss": -83169099.776,
"step": 89000
},
{
"epoch": 13.0,
"eval_accuracy": 0.9390846650576595,
"eval_loss": -83237976.0,
"eval_runtime": 140.4993,
"eval_samples_per_second": 32.562,
"eval_steps_per_second": 5.431,
"step": 89206
},
{
"epoch": 13.04,
"learning_rate": 3.478577674147479e-05,
"loss": -83730743.296,
"step": 89500
},
{
"epoch": 13.12,
"learning_rate": 3.442145147187409e-05,
"loss": -83659464.704,
"step": 90000
},
{
"epoch": 13.19,
"learning_rate": 3.405712620227339e-05,
"loss": -83511828.48,
"step": 90500
},
{
"epoch": 13.26,
"learning_rate": 3.369280093267269e-05,
"loss": -84292861.952,
"step": 91000
},
{
"epoch": 13.33,
"learning_rate": 3.332847566307199e-05,
"loss": -83113410.56,
"step": 91500
},
{
"epoch": 13.41,
"learning_rate": 3.2964150393471295e-05,
"loss": -84954865.664,
"step": 92000
},
{
"epoch": 13.48,
"learning_rate": 3.2599825123870594e-05,
"loss": -84344479.744,
"step": 92500
},
{
"epoch": 13.55,
"learning_rate": 3.22354998542699e-05,
"loss": -84357537.792,
"step": 93000
},
{
"epoch": 13.63,
"learning_rate": 3.18711745846692e-05,
"loss": -83578437.632,
"step": 93500
},
{
"epoch": 13.7,
"learning_rate": 3.1506849315068496e-05,
"loss": -86547087.36,
"step": 94000
},
{
"epoch": 13.77,
"learning_rate": 3.1142524045467795e-05,
"loss": -82968248.32,
"step": 94500
},
{
"epoch": 13.84,
"learning_rate": 3.077819877586709e-05,
"loss": -84810694.656,
"step": 95000
},
{
"epoch": 13.92,
"learning_rate": 3.0413873506266395e-05,
"loss": -86185246.72,
"step": 95500
},
{
"epoch": 13.99,
"learning_rate": 3.0049548236665697e-05,
"loss": -84793049.088,
"step": 96000
},
{
"epoch": 14.0,
"eval_accuracy": 0.9357940716161526,
"eval_loss": -84842016.0,
"eval_runtime": 140.5568,
"eval_samples_per_second": 32.549,
"eval_steps_per_second": 5.428,
"step": 96068
},
{
"epoch": 14.06,
"learning_rate": 2.9685222967064995e-05,
"loss": -84520574.976,
"step": 96500
},
{
"epoch": 14.14,
"learning_rate": 2.93208976974643e-05,
"loss": -85894602.752,
"step": 97000
},
{
"epoch": 14.21,
"learning_rate": 2.89565724278636e-05,
"loss": -85534507.008,
"step": 97500
},
{
"epoch": 14.28,
"learning_rate": 2.8592247158262898e-05,
"loss": -85865168.896,
"step": 98000
},
{
"epoch": 14.35,
"learning_rate": 2.82279218886622e-05,
"loss": -85208301.568,
"step": 98500
},
{
"epoch": 14.43,
"learning_rate": 2.7863596619061498e-05,
"loss": -85955837.952,
"step": 99000
},
{
"epoch": 14.5,
"learning_rate": 2.7499271349460797e-05,
"loss": -86373990.4,
"step": 99500
},
{
"epoch": 14.57,
"learning_rate": 2.7134946079860102e-05,
"loss": -86160285.696,
"step": 100000
},
{
"epoch": 14.65,
"learning_rate": 2.67706208102594e-05,
"loss": -84279885.824,
"step": 100500
},
{
"epoch": 14.72,
"learning_rate": 2.64062955406587e-05,
"loss": -86010617.856,
"step": 101000
},
{
"epoch": 14.79,
"learning_rate": 2.6041970271058004e-05,
"loss": -85548597.248,
"step": 101500
},
{
"epoch": 14.86,
"learning_rate": 2.5677645001457303e-05,
"loss": -85261148.16,
"step": 102000
},
{
"epoch": 14.94,
"learning_rate": 2.5313319731856604e-05,
"loss": -85792735.232,
"step": 102500
},
{
"epoch": 15.0,
"eval_accuracy": 0.9335088520950316,
"eval_loss": -86251000.0,
"eval_runtime": 140.5576,
"eval_samples_per_second": 32.549,
"eval_steps_per_second": 5.428,
"step": 102930
},
{
"epoch": 15.01,
"learning_rate": 2.4948994462255903e-05,
"loss": -86945275.904,
"step": 103000
},
{
"epoch": 15.08,
"learning_rate": 2.4584669192655205e-05,
"loss": -86477979.648,
"step": 103500
},
{
"epoch": 15.16,
"learning_rate": 2.4220343923054503e-05,
"loss": -86069059.584,
"step": 104000
},
{
"epoch": 15.23,
"learning_rate": 2.3856018653453805e-05,
"loss": -87284703.232,
"step": 104500
},
{
"epoch": 15.3,
"learning_rate": 2.3491693383853107e-05,
"loss": -88140079.104,
"step": 105000
},
{
"epoch": 15.37,
"learning_rate": 2.3127368114252406e-05,
"loss": -86128009.216,
"step": 105500
},
{
"epoch": 15.45,
"learning_rate": 2.2763042844651704e-05,
"loss": -87689273.344,
"step": 106000
},
{
"epoch": 15.52,
"learning_rate": 2.2398717575051006e-05,
"loss": -86607101.952,
"step": 106500
},
{
"epoch": 15.59,
"learning_rate": 2.2034392305450308e-05,
"loss": -86783418.368,
"step": 107000
},
{
"epoch": 15.67,
"learning_rate": 2.167006703584961e-05,
"loss": -87358513.152,
"step": 107500
},
{
"epoch": 15.74,
"learning_rate": 2.130574176624891e-05,
"loss": -87125909.504,
"step": 108000
},
{
"epoch": 15.81,
"learning_rate": 2.0941416496648207e-05,
"loss": -86079799.296,
"step": 108500
},
{
"epoch": 15.88,
"learning_rate": 2.057709122704751e-05,
"loss": -86173089.792,
"step": 109000
},
{
"epoch": 15.96,
"learning_rate": 2.0212765957446807e-05,
"loss": -86555295.744,
"step": 109500
},
{
"epoch": 16.0,
"eval_accuracy": 0.7564521289700002,
"eval_loss": -87419264.0,
"eval_runtime": 140.5719,
"eval_samples_per_second": 32.546,
"eval_steps_per_second": 5.428,
"step": 109792
},
{
"epoch": 16.03,
"learning_rate": 1.984844068784611e-05,
"loss": -87372087.296,
"step": 110000
},
{
"epoch": 16.1,
"learning_rate": 1.948411541824541e-05,
"loss": -87514480.64,
"step": 110500
},
{
"epoch": 16.18,
"learning_rate": 1.9119790148644713e-05,
"loss": -86731694.08,
"step": 111000
},
{
"epoch": 16.25,
"learning_rate": 1.875546487904401e-05,
"loss": -88692531.2,
"step": 111500
},
{
"epoch": 16.32,
"learning_rate": 1.839113960944331e-05,
"loss": -87868801.024,
"step": 112000
},
{
"epoch": 16.39,
"learning_rate": 1.8026814339842612e-05,
"loss": -86980747.264,
"step": 112500
},
{
"epoch": 16.47,
"learning_rate": 1.7662489070241914e-05,
"loss": -87147741.184,
"step": 113000
},
{
"epoch": 16.54,
"learning_rate": 1.7298163800641212e-05,
"loss": -87115055.104,
"step": 113500
},
{
"epoch": 16.61,
"learning_rate": 1.6933838531040514e-05,
"loss": -86399164.416,
"step": 114000
},
{
"epoch": 16.69,
"learning_rate": 1.6569513261439816e-05,
"loss": -88488140.8,
"step": 114500
},
{
"epoch": 16.76,
"learning_rate": 1.6205187991839114e-05,
"loss": -87634550.784,
"step": 115000
},
{
"epoch": 16.83,
"learning_rate": 1.5840862722238413e-05,
"loss": -89087279.104,
"step": 115500
},
{
"epoch": 16.9,
"learning_rate": 1.5476537452637715e-05,
"loss": -87880040.448,
"step": 116000
},
{
"epoch": 16.98,
"learning_rate": 1.5112212183037017e-05,
"loss": -88267481.088,
"step": 116500
},
{
"epoch": 17.0,
"eval_accuracy": 0.8551131978455251,
"eval_loss": -88325912.0,
"eval_runtime": 140.5657,
"eval_samples_per_second": 32.547,
"eval_steps_per_second": 5.428,
"step": 116654
},
{
"epoch": 17.05,
"learning_rate": 1.4747886913436319e-05,
"loss": -89539575.808,
"step": 117000
},
{
"epoch": 17.12,
"learning_rate": 1.4383561643835617e-05,
"loss": -88498511.872,
"step": 117500
},
{
"epoch": 17.2,
"learning_rate": 1.4019236374234917e-05,
"loss": -89055494.144,
"step": 118000
},
{
"epoch": 17.27,
"learning_rate": 1.365491110463422e-05,
"loss": -88475312.128,
"step": 118500
},
{
"epoch": 17.34,
"learning_rate": 1.3290585835033518e-05,
"loss": -89511239.68,
"step": 119000
},
{
"epoch": 17.41,
"learning_rate": 1.2926260565432818e-05,
"loss": -88222310.4,
"step": 119500
},
{
"epoch": 17.49,
"learning_rate": 1.256193529583212e-05,
"loss": -87449542.656,
"step": 120000
},
{
"epoch": 17.56,
"learning_rate": 1.219761002623142e-05,
"loss": -88003665.92,
"step": 120500
},
{
"epoch": 17.63,
"learning_rate": 1.183328475663072e-05,
"loss": -88902852.608,
"step": 121000
},
{
"epoch": 17.71,
"learning_rate": 1.146895948703002e-05,
"loss": -89969811.456,
"step": 121500
},
{
"epoch": 17.78,
"learning_rate": 1.1104634217429322e-05,
"loss": -88997732.352,
"step": 122000
},
{
"epoch": 17.85,
"learning_rate": 1.074030894782862e-05,
"loss": -88231747.584,
"step": 122500
},
{
"epoch": 17.92,
"learning_rate": 1.0375983678227923e-05,
"loss": -87653048.32,
"step": 123000
},
{
"epoch": 18.0,
"learning_rate": 1.0011658408627223e-05,
"loss": -87671209.984,
"step": 123500
},
{
"epoch": 18.0,
"eval_accuracy": 0.9378666650310812,
"eval_loss": -88980144.0,
"eval_runtime": 140.5787,
"eval_samples_per_second": 32.544,
"eval_steps_per_second": 5.428,
"step": 123516
},
{
"epoch": 18.07,
"learning_rate": 9.647333139026523e-06,
"loss": -89122250.752,
"step": 124000
},
{
"epoch": 18.14,
"learning_rate": 9.283007869425823e-06,
"loss": -88477589.504,
"step": 124500
},
{
"epoch": 18.22,
"learning_rate": 8.918682599825125e-06,
"loss": -89125781.504,
"step": 125000
},
{
"epoch": 18.29,
"learning_rate": 8.554357330224424e-06,
"loss": -89081135.104,
"step": 125500
},
{
"epoch": 18.36,
"learning_rate": 8.190032060623726e-06,
"loss": -89223471.104,
"step": 126000
},
{
"epoch": 18.43,
"learning_rate": 7.825706791023026e-06,
"loss": -89985130.496,
"step": 126500
},
{
"epoch": 18.51,
"learning_rate": 7.461381521422327e-06,
"loss": -91432869.888,
"step": 127000
},
{
"epoch": 18.58,
"learning_rate": 7.097056251821626e-06,
"loss": -91106402.304,
"step": 127500
},
{
"epoch": 18.65,
"learning_rate": 6.732730982220928e-06,
"loss": -88089059.328,
"step": 128000
},
{
"epoch": 18.73,
"learning_rate": 6.3684057126202274e-06,
"loss": -89137561.6,
"step": 128500
},
{
"epoch": 18.8,
"learning_rate": 6.0040804430195285e-06,
"loss": -88409800.704,
"step": 129000
},
{
"epoch": 18.87,
"learning_rate": 5.639755173418829e-06,
"loss": -89817399.296,
"step": 129500
},
{
"epoch": 18.94,
"learning_rate": 5.275429903818129e-06,
"loss": -90569072.64,
"step": 130000
},
{
"epoch": 19.0,
"eval_accuracy": 0.9353918198239087,
"eval_loss": -89374976.0,
"eval_runtime": 140.5596,
"eval_samples_per_second": 32.548,
"eval_steps_per_second": 5.428,
"step": 130378
}
],
"logging_steps": 500,
"max_steps": 137240,
"num_input_tokens_seen": 0,
"num_train_epochs": 20,
"save_steps": 500,
"total_flos": 5.220527123197805e+17,
"train_batch_size": 6,
"trial_name": null,
"trial_params": null
}