tmnam20's picture
Training in progress, step 48000, checkpoint
95a57da
{
"best_metric": 0.9729449152946472,
"best_model_checkpoint": "/kaggle/output/checkpoint-47000",
"epoch": 1.9556714471968708,
"eval_steps": 1000,
"global_step": 48000,
"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": 1.029,
"step": 1
},
{
"epoch": 0.04,
"learning_rate": 2.7638888888888893e-08,
"loss": 1.181,
"step": 1000
},
{
"epoch": 0.04,
"eval_accuracy": 0.3273453093812375,
"eval_loss": 1.1529844999313354,
"eval_runtime": 54.2837,
"eval_samples_per_second": 92.293,
"eval_steps_per_second": 11.55,
"step": 1000
},
{
"epoch": 0.08,
"learning_rate": 5.541666666666667e-08,
"loss": 1.1527,
"step": 2000
},
{
"epoch": 0.08,
"eval_accuracy": 0.33013972055888224,
"eval_loss": 1.1351025104522705,
"eval_runtime": 54.0518,
"eval_samples_per_second": 92.689,
"eval_steps_per_second": 11.6,
"step": 2000
},
{
"epoch": 0.12,
"learning_rate": 8.316666666666666e-08,
"loss": 1.142,
"step": 3000
},
{
"epoch": 0.12,
"eval_accuracy": 0.3317365269461078,
"eval_loss": 1.127414345741272,
"eval_runtime": 54.0871,
"eval_samples_per_second": 92.628,
"eval_steps_per_second": 11.592,
"step": 3000
},
{
"epoch": 0.16,
"learning_rate": 1.1091666666666668e-07,
"loss": 1.1371,
"step": 4000
},
{
"epoch": 0.16,
"eval_accuracy": 0.331936127744511,
"eval_loss": 1.1194497346878052,
"eval_runtime": 54.3907,
"eval_samples_per_second": 92.111,
"eval_steps_per_second": 11.528,
"step": 4000
},
{
"epoch": 0.2,
"learning_rate": 1.3869444444444447e-07,
"loss": 1.1246,
"step": 5000
},
{
"epoch": 0.2,
"eval_accuracy": 0.3345309381237525,
"eval_loss": 1.1131775379180908,
"eval_runtime": 54.3492,
"eval_samples_per_second": 92.182,
"eval_steps_per_second": 11.537,
"step": 5000
},
{
"epoch": 0.24,
"learning_rate": 1.664722222222222e-07,
"loss": 1.1201,
"step": 6000
},
{
"epoch": 0.24,
"eval_accuracy": 0.34151696606786425,
"eval_loss": 1.1093512773513794,
"eval_runtime": 54.3536,
"eval_samples_per_second": 92.174,
"eval_steps_per_second": 11.536,
"step": 6000
},
{
"epoch": 0.29,
"learning_rate": 1.9425e-07,
"loss": 1.115,
"step": 7000
},
{
"epoch": 0.29,
"eval_accuracy": 0.34510978043912177,
"eval_loss": 1.1066845655441284,
"eval_runtime": 53.9977,
"eval_samples_per_second": 92.782,
"eval_steps_per_second": 11.612,
"step": 7000
},
{
"epoch": 0.33,
"learning_rate": 2.2202777777777778e-07,
"loss": 1.1125,
"step": 8000
},
{
"epoch": 0.33,
"eval_accuracy": 0.3449101796407186,
"eval_loss": 1.1031057834625244,
"eval_runtime": 54.2562,
"eval_samples_per_second": 92.34,
"eval_steps_per_second": 11.556,
"step": 8000
},
{
"epoch": 0.37,
"learning_rate": 2.498055555555556e-07,
"loss": 1.1072,
"step": 9000
},
{
"epoch": 0.37,
"eval_accuracy": 0.34870259481037924,
"eval_loss": 1.1044224500656128,
"eval_runtime": 54.0795,
"eval_samples_per_second": 92.641,
"eval_steps_per_second": 11.594,
"step": 9000
},
{
"epoch": 0.41,
"learning_rate": 2.7755555555555554e-07,
"loss": 1.1064,
"step": 10000
},
{
"epoch": 0.41,
"eval_accuracy": 0.3499001996007984,
"eval_loss": 1.1001031398773193,
"eval_runtime": 54.0869,
"eval_samples_per_second": 92.629,
"eval_steps_per_second": 11.592,
"step": 10000
},
{
"epoch": 0.45,
"learning_rate": 3.053333333333333e-07,
"loss": 1.1053,
"step": 11000
},
{
"epoch": 0.45,
"eval_accuracy": 0.36087824351297404,
"eval_loss": 1.0987293720245361,
"eval_runtime": 53.9688,
"eval_samples_per_second": 92.831,
"eval_steps_per_second": 11.618,
"step": 11000
},
{
"epoch": 0.49,
"learning_rate": 3.3311111111111114e-07,
"loss": 1.104,
"step": 12000
},
{
"epoch": 0.49,
"eval_accuracy": 0.3594810379241517,
"eval_loss": 1.0968631505966187,
"eval_runtime": 54.0936,
"eval_samples_per_second": 92.617,
"eval_steps_per_second": 11.591,
"step": 12000
},
{
"epoch": 0.53,
"learning_rate": 3.608888888888889e-07,
"loss": 1.1022,
"step": 13000
},
{
"epoch": 0.53,
"eval_accuracy": 0.36087824351297404,
"eval_loss": 1.0962979793548584,
"eval_runtime": 54.0403,
"eval_samples_per_second": 92.709,
"eval_steps_per_second": 11.602,
"step": 13000
},
{
"epoch": 0.57,
"learning_rate": 3.886666666666667e-07,
"loss": 1.1029,
"step": 14000
},
{
"epoch": 0.57,
"eval_accuracy": 0.3528942115768463,
"eval_loss": 1.0987964868545532,
"eval_runtime": 54.06,
"eval_samples_per_second": 92.675,
"eval_steps_per_second": 11.598,
"step": 14000
},
{
"epoch": 0.61,
"learning_rate": 4.163611111111111e-07,
"loss": 1.1011,
"step": 15000
},
{
"epoch": 0.61,
"eval_accuracy": 0.36926147704590817,
"eval_loss": 1.0941977500915527,
"eval_runtime": 54.0807,
"eval_samples_per_second": 92.639,
"eval_steps_per_second": 11.594,
"step": 15000
},
{
"epoch": 0.65,
"learning_rate": 4.441388888888889e-07,
"loss": 1.1004,
"step": 16000
},
{
"epoch": 0.65,
"eval_accuracy": 0.36866267465069863,
"eval_loss": 1.093828558921814,
"eval_runtime": 54.2208,
"eval_samples_per_second": 92.4,
"eval_steps_per_second": 11.564,
"step": 16000
},
{
"epoch": 0.69,
"learning_rate": 4.719166666666667e-07,
"loss": 1.0994,
"step": 17000
},
{
"epoch": 0.69,
"eval_accuracy": 0.36726546906187624,
"eval_loss": 1.093470811843872,
"eval_runtime": 54.2584,
"eval_samples_per_second": 92.336,
"eval_steps_per_second": 11.556,
"step": 17000
},
{
"epoch": 0.73,
"learning_rate": 4.996944444444445e-07,
"loss": 1.0993,
"step": 18000
},
{
"epoch": 0.73,
"eval_accuracy": 0.36327345309381237,
"eval_loss": 1.0935499668121338,
"eval_runtime": 54.0554,
"eval_samples_per_second": 92.683,
"eval_steps_per_second": 11.599,
"step": 18000
},
{
"epoch": 0.77,
"learning_rate": 5.274722222222222e-07,
"loss": 1.0981,
"step": 19000
},
{
"epoch": 0.77,
"eval_accuracy": 0.38682634730538923,
"eval_loss": 1.089295506477356,
"eval_runtime": 54.0761,
"eval_samples_per_second": 92.647,
"eval_steps_per_second": 11.595,
"step": 19000
},
{
"epoch": 0.81,
"learning_rate": 5.5525e-07,
"loss": 1.0979,
"step": 20000
},
{
"epoch": 0.81,
"eval_accuracy": 0.4167664670658683,
"eval_loss": 1.084443211555481,
"eval_runtime": 54.0394,
"eval_samples_per_second": 92.71,
"eval_steps_per_second": 11.603,
"step": 20000
},
{
"epoch": 0.86,
"learning_rate": 5.830000000000001e-07,
"loss": 1.0946,
"step": 21000
},
{
"epoch": 0.86,
"eval_accuracy": 0.4285429141716567,
"eval_loss": 1.073406457901001,
"eval_runtime": 54.1561,
"eval_samples_per_second": 92.51,
"eval_steps_per_second": 11.578,
"step": 21000
},
{
"epoch": 0.9,
"learning_rate": 6.107777777777778e-07,
"loss": 1.0882,
"step": 22000
},
{
"epoch": 0.9,
"eval_accuracy": 0.4413173652694611,
"eval_loss": 1.0576634407043457,
"eval_runtime": 53.9715,
"eval_samples_per_second": 92.827,
"eval_steps_per_second": 11.617,
"step": 22000
},
{
"epoch": 0.94,
"learning_rate": 6.385277777777778e-07,
"loss": 1.0811,
"step": 23000
},
{
"epoch": 0.94,
"eval_accuracy": 0.4377245508982036,
"eval_loss": 1.063384771347046,
"eval_runtime": 54.4655,
"eval_samples_per_second": 91.985,
"eval_steps_per_second": 11.512,
"step": 23000
},
{
"epoch": 0.98,
"learning_rate": 6.662777777777778e-07,
"loss": 1.0773,
"step": 24000
},
{
"epoch": 0.98,
"eval_accuracy": 0.4347305389221557,
"eval_loss": 1.065600037574768,
"eval_runtime": 54.0545,
"eval_samples_per_second": 92.684,
"eval_steps_per_second": 11.599,
"step": 24000
},
{
"epoch": 1.02,
"learning_rate": 6.940555555555557e-07,
"loss": 1.0749,
"step": 25000
},
{
"epoch": 1.02,
"eval_accuracy": 0.4499001996007984,
"eval_loss": 1.0491468906402588,
"eval_runtime": 53.9397,
"eval_samples_per_second": 92.882,
"eval_steps_per_second": 11.624,
"step": 25000
},
{
"epoch": 1.06,
"learning_rate": 7.218055555555556e-07,
"loss": 1.0687,
"step": 26000
},
{
"epoch": 1.06,
"eval_accuracy": 0.44471057884231535,
"eval_loss": 1.0531450510025024,
"eval_runtime": 53.996,
"eval_samples_per_second": 92.785,
"eval_steps_per_second": 11.612,
"step": 26000
},
{
"epoch": 1.1,
"learning_rate": 7.495833333333334e-07,
"loss": 1.07,
"step": 27000
},
{
"epoch": 1.1,
"eval_accuracy": 0.4475049900199601,
"eval_loss": 1.0514291524887085,
"eval_runtime": 54.0862,
"eval_samples_per_second": 92.63,
"eval_steps_per_second": 11.593,
"step": 27000
},
{
"epoch": 1.14,
"learning_rate": 7.773611111111112e-07,
"loss": 1.0619,
"step": 28000
},
{
"epoch": 1.14,
"eval_accuracy": 0.4483033932135729,
"eval_loss": 1.0422637462615967,
"eval_runtime": 54.3061,
"eval_samples_per_second": 92.255,
"eval_steps_per_second": 11.546,
"step": 28000
},
{
"epoch": 1.18,
"learning_rate": 8.051388888888889e-07,
"loss": 1.0586,
"step": 29000
},
{
"epoch": 1.18,
"eval_accuracy": 0.4544910179640719,
"eval_loss": 1.0364656448364258,
"eval_runtime": 54.1221,
"eval_samples_per_second": 92.568,
"eval_steps_per_second": 11.585,
"step": 29000
},
{
"epoch": 1.22,
"learning_rate": 8.32888888888889e-07,
"loss": 1.0571,
"step": 30000
},
{
"epoch": 1.22,
"eval_accuracy": 0.4594810379241517,
"eval_loss": 1.0300606489181519,
"eval_runtime": 54.1093,
"eval_samples_per_second": 92.59,
"eval_steps_per_second": 11.588,
"step": 30000
},
{
"epoch": 1.26,
"learning_rate": 8.606666666666668e-07,
"loss": 1.0564,
"step": 31000
},
{
"epoch": 1.26,
"eval_accuracy": 0.4666666666666667,
"eval_loss": 1.0317378044128418,
"eval_runtime": 54.0415,
"eval_samples_per_second": 92.707,
"eval_steps_per_second": 11.602,
"step": 31000
},
{
"epoch": 1.3,
"learning_rate": 8.884444444444445e-07,
"loss": 1.0541,
"step": 32000
},
{
"epoch": 1.3,
"eval_accuracy": 0.47065868263473054,
"eval_loss": 1.0293121337890625,
"eval_runtime": 54.1157,
"eval_samples_per_second": 92.579,
"eval_steps_per_second": 11.586,
"step": 32000
},
{
"epoch": 1.34,
"learning_rate": 9.161944444444445e-07,
"loss": 1.0485,
"step": 33000
},
{
"epoch": 1.34,
"eval_accuracy": 0.480439121756487,
"eval_loss": 1.0269454717636108,
"eval_runtime": 54.0627,
"eval_samples_per_second": 92.67,
"eval_steps_per_second": 11.598,
"step": 33000
},
{
"epoch": 1.39,
"learning_rate": 9.439722222222223e-07,
"loss": 1.0469,
"step": 34000
},
{
"epoch": 1.39,
"eval_accuracy": 0.4780439121756487,
"eval_loss": 1.026614785194397,
"eval_runtime": 54.0605,
"eval_samples_per_second": 92.674,
"eval_steps_per_second": 11.598,
"step": 34000
},
{
"epoch": 1.43,
"learning_rate": 9.7175e-07,
"loss": 1.0479,
"step": 35000
},
{
"epoch": 1.43,
"eval_accuracy": 0.48423153692614773,
"eval_loss": 1.0213700532913208,
"eval_runtime": 54.1341,
"eval_samples_per_second": 92.548,
"eval_steps_per_second": 11.582,
"step": 35000
},
{
"epoch": 1.47,
"learning_rate": 9.995277777777778e-07,
"loss": 1.0433,
"step": 36000
},
{
"epoch": 1.47,
"eval_accuracy": 0.49101796407185627,
"eval_loss": 1.0118540525436401,
"eval_runtime": 54.1201,
"eval_samples_per_second": 92.572,
"eval_steps_per_second": 11.585,
"step": 36000
},
{
"epoch": 1.51,
"learning_rate": 1.0273055555555556e-06,
"loss": 1.0368,
"step": 37000
},
{
"epoch": 1.51,
"eval_accuracy": 0.49101796407185627,
"eval_loss": 1.0219311714172363,
"eval_runtime": 54.1135,
"eval_samples_per_second": 92.583,
"eval_steps_per_second": 11.587,
"step": 37000
},
{
"epoch": 1.55,
"learning_rate": 1.0550555555555557e-06,
"loss": 1.0401,
"step": 38000
},
{
"epoch": 1.55,
"eval_accuracy": 0.49720558882235527,
"eval_loss": 1.013930320739746,
"eval_runtime": 54.0309,
"eval_samples_per_second": 92.725,
"eval_steps_per_second": 11.604,
"step": 38000
},
{
"epoch": 1.59,
"learning_rate": 1.0828333333333334e-06,
"loss": 1.0341,
"step": 39000
},
{
"epoch": 1.59,
"eval_accuracy": 0.4992015968063872,
"eval_loss": 1.0132442712783813,
"eval_runtime": 54.0478,
"eval_samples_per_second": 92.696,
"eval_steps_per_second": 11.601,
"step": 39000
},
{
"epoch": 1.63,
"learning_rate": 1.1106111111111112e-06,
"loss": 1.0254,
"step": 40000
},
{
"epoch": 1.63,
"eval_accuracy": 0.5095808383233533,
"eval_loss": 1.0034514665603638,
"eval_runtime": 54.1138,
"eval_samples_per_second": 92.583,
"eval_steps_per_second": 11.587,
"step": 40000
},
{
"epoch": 1.67,
"learning_rate": 1.1383611111111113e-06,
"loss": 1.0264,
"step": 41000
},
{
"epoch": 1.67,
"eval_accuracy": 0.5159680638722555,
"eval_loss": 1.0014362335205078,
"eval_runtime": 54.2855,
"eval_samples_per_second": 92.29,
"eval_steps_per_second": 11.55,
"step": 41000
},
{
"epoch": 1.71,
"learning_rate": 1.166138888888889e-06,
"loss": 1.0222,
"step": 42000
},
{
"epoch": 1.71,
"eval_accuracy": 0.5183632734530939,
"eval_loss": 0.9864664077758789,
"eval_runtime": 54.1055,
"eval_samples_per_second": 92.597,
"eval_steps_per_second": 11.588,
"step": 42000
},
{
"epoch": 1.75,
"learning_rate": 1.193888888888889e-06,
"loss": 1.017,
"step": 43000
},
{
"epoch": 1.75,
"eval_accuracy": 0.5153692614770459,
"eval_loss": 0.995494544506073,
"eval_runtime": 54.2838,
"eval_samples_per_second": 92.293,
"eval_steps_per_second": 11.55,
"step": 43000
},
{
"epoch": 1.79,
"learning_rate": 1.2216666666666667e-06,
"loss": 1.0225,
"step": 44000
},
{
"epoch": 1.79,
"eval_accuracy": 0.5165668662674651,
"eval_loss": 0.9881771802902222,
"eval_runtime": 54.6114,
"eval_samples_per_second": 91.739,
"eval_steps_per_second": 11.481,
"step": 44000
},
{
"epoch": 1.83,
"learning_rate": 1.2494444444444447e-06,
"loss": 1.0149,
"step": 45000
},
{
"epoch": 1.83,
"eval_accuracy": 0.5265469061876248,
"eval_loss": 0.9803952574729919,
"eval_runtime": 54.6026,
"eval_samples_per_second": 91.754,
"eval_steps_per_second": 11.483,
"step": 45000
},
{
"epoch": 1.87,
"learning_rate": 1.2772222222222224e-06,
"loss": 1.017,
"step": 46000
},
{
"epoch": 1.87,
"eval_accuracy": 0.5331337325349301,
"eval_loss": 0.9766710996627808,
"eval_runtime": 54.449,
"eval_samples_per_second": 92.013,
"eval_steps_per_second": 11.515,
"step": 46000
},
{
"epoch": 1.91,
"learning_rate": 1.3049722222222223e-06,
"loss": 1.0012,
"step": 47000
},
{
"epoch": 1.91,
"eval_accuracy": 0.5317365269461077,
"eval_loss": 0.9729449152946472,
"eval_runtime": 54.1692,
"eval_samples_per_second": 92.488,
"eval_steps_per_second": 11.575,
"step": 47000
},
{
"epoch": 1.96,
"learning_rate": 1.3327222222222224e-06,
"loss": 1.0039,
"step": 48000
},
{
"epoch": 1.96,
"eval_accuracy": 0.5259481037924152,
"eval_loss": 0.9904573559761047,
"eval_runtime": 54.3246,
"eval_samples_per_second": 92.223,
"eval_steps_per_second": 11.542,
"step": 48000
}
],
"logging_steps": 1000,
"max_steps": 10000000,
"num_train_epochs": 408,
"save_steps": 1000,
"total_flos": 1.003367849632727e+17,
"trial_name": null,
"trial_params": null
}