{
"best_metric": 1.0493717193603516,
"best_model_checkpoint": "/kaggle/output/checkpoint-30000",
"epoch": 1.303780964797914,
"eval_steps": 1000,
"global_step": 32000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 2.7777777777777777e-11,
"loss": 1.2184,
"step": 1
},
{
"epoch": 0.04,
"learning_rate": 2.7750000000000004e-08,
"loss": 1.1394,
"step": 1000
},
{
"epoch": 0.04,
"eval_accuracy": 0.3327345309381238,
"eval_loss": 1.1149410009384155,
"eval_runtime": 20.6803,
"eval_samples_per_second": 242.26,
"eval_steps_per_second": 30.319,
"step": 1000
},
{
"epoch": 0.08,
"learning_rate": 5.5527777777777784e-08,
"loss": 1.1141,
"step": 2000
},
{
"epoch": 0.08,
"eval_accuracy": 0.3401197604790419,
"eval_loss": 1.104099988937378,
"eval_runtime": 20.8477,
"eval_samples_per_second": 240.314,
"eval_steps_per_second": 30.075,
"step": 2000
},
{
"epoch": 0.12,
"learning_rate": 8.330555555555556e-08,
"loss": 1.116,
"step": 3000
},
{
"epoch": 0.12,
"eval_accuracy": 0.3407185628742515,
"eval_loss": 1.1040862798690796,
"eval_runtime": 20.6818,
"eval_samples_per_second": 242.242,
"eval_steps_per_second": 30.317,
"step": 3000
},
{
"epoch": 0.16,
"learning_rate": 1.1108333333333333e-07,
"loss": 1.1158,
"step": 4000
},
{
"epoch": 0.16,
"eval_accuracy": 0.32894211576846305,
"eval_loss": 1.1020556688308716,
"eval_runtime": 20.8541,
"eval_samples_per_second": 240.241,
"eval_steps_per_second": 30.066,
"step": 4000
},
{
"epoch": 0.2,
"learning_rate": 1.3883333333333335e-07,
"loss": 1.1135,
"step": 5000
},
{
"epoch": 0.2,
"eval_accuracy": 0.34271457085828344,
"eval_loss": 1.1008552312850952,
"eval_runtime": 20.8055,
"eval_samples_per_second": 240.802,
"eval_steps_per_second": 30.136,
"step": 5000
},
{
"epoch": 0.24,
"learning_rate": 1.6658333333333335e-07,
"loss": 1.1121,
"step": 6000
},
{
"epoch": 0.24,
"eval_accuracy": 0.3395209580838323,
"eval_loss": 1.1004050970077515,
"eval_runtime": 20.8985,
"eval_samples_per_second": 239.731,
"eval_steps_per_second": 30.002,
"step": 6000
},
{
"epoch": 0.29,
"learning_rate": 1.9436111111111112e-07,
"loss": 1.1089,
"step": 7000
},
{
"epoch": 0.29,
"eval_accuracy": 0.35788423153692617,
"eval_loss": 1.0985721349716187,
"eval_runtime": 20.84,
"eval_samples_per_second": 240.403,
"eval_steps_per_second": 30.086,
"step": 7000
},
{
"epoch": 0.33,
"learning_rate": 2.2213888888888891e-07,
"loss": 1.1079,
"step": 8000
},
{
"epoch": 0.33,
"eval_accuracy": 0.3331337325349301,
"eval_loss": 1.098374843597412,
"eval_runtime": 20.7886,
"eval_samples_per_second": 240.998,
"eval_steps_per_second": 30.161,
"step": 8000
},
{
"epoch": 0.37,
"learning_rate": 2.4988888888888893e-07,
"loss": 1.1087,
"step": 9000
},
{
"epoch": 0.37,
"eval_accuracy": 0.34510978043912177,
"eval_loss": 1.0993521213531494,
"eval_runtime": 20.782,
"eval_samples_per_second": 241.074,
"eval_steps_per_second": 30.17,
"step": 9000
},
{
"epoch": 0.41,
"learning_rate": 2.776666666666667e-07,
"loss": 1.109,
"step": 10000
},
{
"epoch": 0.41,
"eval_accuracy": 0.3475049900199601,
"eval_loss": 1.0967597961425781,
"eval_runtime": 20.6798,
"eval_samples_per_second": 242.265,
"eval_steps_per_second": 30.319,
"step": 10000
},
{
"epoch": 0.45,
"learning_rate": 3.054444444444444e-07,
"loss": 1.1052,
"step": 11000
},
{
"epoch": 0.45,
"eval_accuracy": 0.37544910179640717,
"eval_loss": 1.0941349267959595,
"eval_runtime": 20.8641,
"eval_samples_per_second": 240.126,
"eval_steps_per_second": 30.052,
"step": 11000
},
{
"epoch": 0.49,
"learning_rate": 3.3322222222222225e-07,
"loss": 1.105,
"step": 12000
},
{
"epoch": 0.49,
"eval_accuracy": 0.3834331337325349,
"eval_loss": 1.0927647352218628,
"eval_runtime": 20.6541,
"eval_samples_per_second": 242.567,
"eval_steps_per_second": 30.357,
"step": 12000
},
{
"epoch": 0.53,
"learning_rate": 3.609722222222222e-07,
"loss": 1.1016,
"step": 13000
},
{
"epoch": 0.53,
"eval_accuracy": 0.3457085828343313,
"eval_loss": 1.0942081212997437,
"eval_runtime": 21.0733,
"eval_samples_per_second": 237.742,
"eval_steps_per_second": 29.753,
"step": 13000
},
{
"epoch": 0.57,
"learning_rate": 3.8875e-07,
"loss": 1.1031,
"step": 14000
},
{
"epoch": 0.57,
"eval_accuracy": 0.37005988023952097,
"eval_loss": 1.0918152332305908,
"eval_runtime": 20.9151,
"eval_samples_per_second": 239.54,
"eval_steps_per_second": 29.978,
"step": 14000
},
{
"epoch": 0.61,
"learning_rate": 4.1652777777777786e-07,
"loss": 1.1026,
"step": 15000
},
{
"epoch": 0.61,
"eval_accuracy": 0.3790419161676647,
"eval_loss": 1.0895211696624756,
"eval_runtime": 21.0591,
"eval_samples_per_second": 237.902,
"eval_steps_per_second": 29.773,
"step": 15000
},
{
"epoch": 0.65,
"learning_rate": 4.4427777777777783e-07,
"loss": 1.0988,
"step": 16000
},
{
"epoch": 0.65,
"eval_accuracy": 0.4101796407185629,
"eval_loss": 1.0852997303009033,
"eval_runtime": 20.9509,
"eval_samples_per_second": 239.131,
"eval_steps_per_second": 29.927,
"step": 16000
},
{
"epoch": 0.69,
"learning_rate": 4.720555555555556e-07,
"loss": 1.0974,
"step": 17000
},
{
"epoch": 0.69,
"eval_accuracy": 0.43213572854291415,
"eval_loss": 1.0791982412338257,
"eval_runtime": 20.7526,
"eval_samples_per_second": 241.415,
"eval_steps_per_second": 30.213,
"step": 17000
},
{
"epoch": 0.73,
"learning_rate": 4.998055555555556e-07,
"loss": 1.0932,
"step": 18000
},
{
"epoch": 0.73,
"eval_accuracy": 0.4275449101796407,
"eval_loss": 1.072191596031189,
"eval_runtime": 21.2435,
"eval_samples_per_second": 235.837,
"eval_steps_per_second": 29.515,
"step": 18000
},
{
"epoch": 0.77,
"learning_rate": 5.275833333333334e-07,
"loss": 1.0833,
"step": 19000
},
{
"epoch": 0.77,
"eval_accuracy": 0.43233532934131735,
"eval_loss": 1.06425940990448,
"eval_runtime": 20.7923,
"eval_samples_per_second": 240.955,
"eval_steps_per_second": 30.155,
"step": 19000
},
{
"epoch": 0.81,
"learning_rate": 5.553333333333334e-07,
"loss": 1.0787,
"step": 20000
},
{
"epoch": 0.81,
"eval_accuracy": 0.4295409181636727,
"eval_loss": 1.0638529062271118,
"eval_runtime": 21.0018,
"eval_samples_per_second": 238.551,
"eval_steps_per_second": 29.855,
"step": 20000
},
{
"epoch": 0.86,
"learning_rate": 5.831111111111111e-07,
"loss": 1.0779,
"step": 21000
},
{
"epoch": 0.86,
"eval_accuracy": 0.4243512974051896,
"eval_loss": 1.0603673458099365,
"eval_runtime": 20.9689,
"eval_samples_per_second": 238.926,
"eval_steps_per_second": 29.901,
"step": 21000
},
{
"epoch": 0.9,
"learning_rate": 6.108888888888888e-07,
"loss": 1.0751,
"step": 22000
},
{
"epoch": 0.9,
"eval_accuracy": 0.43233532934131735,
"eval_loss": 1.0603009462356567,
"eval_runtime": 20.8897,
"eval_samples_per_second": 239.831,
"eval_steps_per_second": 30.015,
"step": 22000
},
{
"epoch": 0.94,
"learning_rate": 6.386388888888889e-07,
"loss": 1.0776,
"step": 23000
},
{
"epoch": 0.94,
"eval_accuracy": 0.42734530938123755,
"eval_loss": 1.0591468811035156,
"eval_runtime": 20.964,
"eval_samples_per_second": 238.981,
"eval_steps_per_second": 29.908,
"step": 23000
},
{
"epoch": 0.98,
"learning_rate": 6.664166666666667e-07,
"loss": 1.0754,
"step": 24000
},
{
"epoch": 0.98,
"eval_accuracy": 0.4245508982035928,
"eval_loss": 1.0589721202850342,
"eval_runtime": 20.9053,
"eval_samples_per_second": 239.652,
"eval_steps_per_second": 29.992,
"step": 24000
},
{
"epoch": 1.02,
"learning_rate": 6.941666666666667e-07,
"loss": 1.0736,
"step": 25000
},
{
"epoch": 1.02,
"eval_accuracy": 0.43213572854291415,
"eval_loss": 1.0583962202072144,
"eval_runtime": 21.3265,
"eval_samples_per_second": 234.919,
"eval_steps_per_second": 29.4,
"step": 25000
},
{
"epoch": 1.06,
"learning_rate": 7.219444444444444e-07,
"loss": 1.0717,
"step": 26000
},
{
"epoch": 1.06,
"eval_accuracy": 0.4305389221556886,
"eval_loss": 1.0561293363571167,
"eval_runtime": 21.3034,
"eval_samples_per_second": 235.174,
"eval_steps_per_second": 29.432,
"step": 26000
},
{
"epoch": 1.1,
"learning_rate": 7.496944444444444e-07,
"loss": 1.0709,
"step": 27000
},
{
"epoch": 1.1,
"eval_accuracy": 0.4281437125748503,
"eval_loss": 1.0555357933044434,
"eval_runtime": 21.2178,
"eval_samples_per_second": 236.123,
"eval_steps_per_second": 29.551,
"step": 27000
},
{
"epoch": 1.14,
"learning_rate": 7.774722222222223e-07,
"loss": 1.0701,
"step": 28000
},
{
"epoch": 1.14,
"eval_accuracy": 0.4217564870259481,
"eval_loss": 1.054961085319519,
"eval_runtime": 21.1775,
"eval_samples_per_second": 236.571,
"eval_steps_per_second": 29.607,
"step": 28000
},
{
"epoch": 1.18,
"learning_rate": 8.052222222222223e-07,
"loss": 1.0641,
"step": 29000
},
{
"epoch": 1.18,
"eval_accuracy": 0.4291417165668663,
"eval_loss": 1.0518379211425781,
"eval_runtime": 21.0932,
"eval_samples_per_second": 237.517,
"eval_steps_per_second": 29.725,
"step": 29000
},
{
"epoch": 1.22,
"learning_rate": 8.330000000000001e-07,
"loss": 1.064,
"step": 30000
},
{
"epoch": 1.22,
"eval_accuracy": 0.43173652694610776,
"eval_loss": 1.0493717193603516,
"eval_runtime": 21.2843,
"eval_samples_per_second": 235.385,
"eval_steps_per_second": 29.458,
"step": 30000
},
{
"epoch": 1.26,
"learning_rate": 8.607500000000001e-07,
"loss": 1.0693,
"step": 31000
},
{
"epoch": 1.26,
"eval_accuracy": 0.4291417165668663,
"eval_loss": 1.0521764755249023,
"eval_runtime": 21.2278,
"eval_samples_per_second": 236.011,
"eval_steps_per_second": 29.537,
"step": 31000
},
{
"epoch": 1.3,
"learning_rate": 8.885277777777779e-07,
"loss": 1.0649,
"step": 32000
},
{
"epoch": 1.3,
"eval_accuracy": 0.42375249500998,
"eval_loss": 1.0528494119644165,
"eval_runtime": 21.1249,
"eval_samples_per_second": 237.161,
"eval_steps_per_second": 29.681,
"step": 32000
}
],
"logging_steps": 1000,
"max_steps": 10000000,
"num_train_epochs": 408,
"save_steps": 1000,
"total_flos": 2.2298242247294976e+16,
"trial_name": null,
"trial_params": null
}