tmnam20's picture
Training in progress, step 53000, checkpoint
4eaf891
{
"best_metric": 0.9917733073234558,
"best_model_checkpoint": "/kaggle/output/checkpoint-53000",
"epoch": 2.159387222946545,
"eval_steps": 1000,
"global_step": 53000,
"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": 1.1506,
"step": 1
},
{
"epoch": 0.04,
"learning_rate": 2.7750000000000004e-08,
"loss": 1.1326,
"step": 1000
},
{
"epoch": 0.04,
"eval_accuracy": 0.3331337325349301,
"eval_loss": 1.1057127714157104,
"eval_runtime": 61.3538,
"eval_samples_per_second": 81.657,
"eval_steps_per_second": 10.219,
"step": 1000
},
{
"epoch": 0.08,
"learning_rate": 5.5527777777777784e-08,
"loss": 1.1238,
"step": 2000
},
{
"epoch": 0.08,
"eval_accuracy": 0.33812375249501,
"eval_loss": 1.1044437885284424,
"eval_runtime": 61.256,
"eval_samples_per_second": 81.788,
"eval_steps_per_second": 10.236,
"step": 2000
},
{
"epoch": 0.12,
"learning_rate": 8.327777777777778e-08,
"loss": 1.1251,
"step": 3000
},
{
"epoch": 0.12,
"eval_accuracy": 0.3407185628742515,
"eval_loss": 1.1043975353240967,
"eval_runtime": 61.2433,
"eval_samples_per_second": 81.805,
"eval_steps_per_second": 10.238,
"step": 3000
},
{
"epoch": 0.16,
"learning_rate": 1.1105555555555557e-07,
"loss": 1.1205,
"step": 4000
},
{
"epoch": 0.16,
"eval_accuracy": 0.3387225548902196,
"eval_loss": 1.100860357284546,
"eval_runtime": 61.4457,
"eval_samples_per_second": 81.535,
"eval_steps_per_second": 10.204,
"step": 4000
},
{
"epoch": 0.2,
"learning_rate": 1.3880555555555558e-07,
"loss": 1.1192,
"step": 5000
},
{
"epoch": 0.2,
"eval_accuracy": 0.34151696606786425,
"eval_loss": 1.1005476713180542,
"eval_runtime": 61.2568,
"eval_samples_per_second": 81.787,
"eval_steps_per_second": 10.236,
"step": 5000
},
{
"epoch": 0.24,
"learning_rate": 1.6658333333333335e-07,
"loss": 1.1145,
"step": 6000
},
{
"epoch": 0.24,
"eval_accuracy": 0.3508982035928144,
"eval_loss": 1.0979759693145752,
"eval_runtime": 61.2221,
"eval_samples_per_second": 81.833,
"eval_steps_per_second": 10.241,
"step": 6000
},
{
"epoch": 0.29,
"learning_rate": 1.9433333333333334e-07,
"loss": 1.1164,
"step": 7000
},
{
"epoch": 0.29,
"eval_accuracy": 0.35508982035928144,
"eval_loss": 1.0955135822296143,
"eval_runtime": 61.255,
"eval_samples_per_second": 81.789,
"eval_steps_per_second": 10.236,
"step": 7000
},
{
"epoch": 0.33,
"learning_rate": 2.2211111111111114e-07,
"loss": 1.1091,
"step": 8000
},
{
"epoch": 0.33,
"eval_accuracy": 0.36147704590818364,
"eval_loss": 1.0946425199508667,
"eval_runtime": 61.3011,
"eval_samples_per_second": 81.728,
"eval_steps_per_second": 10.228,
"step": 8000
},
{
"epoch": 0.37,
"learning_rate": 2.4986111111111113e-07,
"loss": 1.107,
"step": 9000
},
{
"epoch": 0.37,
"eval_accuracy": 0.3471057884231537,
"eval_loss": 1.10226309299469,
"eval_runtime": 61.3193,
"eval_samples_per_second": 81.703,
"eval_steps_per_second": 10.225,
"step": 9000
},
{
"epoch": 0.41,
"learning_rate": 2.776388888888889e-07,
"loss": 1.1055,
"step": 10000
},
{
"epoch": 0.41,
"eval_accuracy": 0.3774451097804391,
"eval_loss": 1.0922847986221313,
"eval_runtime": 61.1866,
"eval_samples_per_second": 81.881,
"eval_steps_per_second": 10.247,
"step": 10000
},
{
"epoch": 0.45,
"learning_rate": 3.0541666666666667e-07,
"loss": 1.1028,
"step": 11000
},
{
"epoch": 0.45,
"eval_accuracy": 0.37425149700598803,
"eval_loss": 1.0914534330368042,
"eval_runtime": 61.341,
"eval_samples_per_second": 81.675,
"eval_steps_per_second": 10.222,
"step": 11000
},
{
"epoch": 0.49,
"learning_rate": 3.3319444444444444e-07,
"loss": 1.1024,
"step": 12000
},
{
"epoch": 0.49,
"eval_accuracy": 0.3872255489021956,
"eval_loss": 1.0888582468032837,
"eval_runtime": 61.4245,
"eval_samples_per_second": 81.564,
"eval_steps_per_second": 10.208,
"step": 12000
},
{
"epoch": 0.53,
"learning_rate": 3.6094444444444446e-07,
"loss": 1.1015,
"step": 13000
},
{
"epoch": 0.53,
"eval_accuracy": 0.3397205588822355,
"eval_loss": 1.096097707748413,
"eval_runtime": 61.3257,
"eval_samples_per_second": 81.695,
"eval_steps_per_second": 10.224,
"step": 13000
},
{
"epoch": 0.57,
"learning_rate": 3.8872222222222223e-07,
"loss": 1.1,
"step": 14000
},
{
"epoch": 0.57,
"eval_accuracy": 0.36387225548902197,
"eval_loss": 1.0940359830856323,
"eval_runtime": 61.2945,
"eval_samples_per_second": 81.737,
"eval_steps_per_second": 10.229,
"step": 14000
},
{
"epoch": 0.61,
"learning_rate": 4.1650000000000006e-07,
"loss": 1.0997,
"step": 15000
},
{
"epoch": 0.61,
"eval_accuracy": 0.38602794411177643,
"eval_loss": 1.0880299806594849,
"eval_runtime": 61.2546,
"eval_samples_per_second": 81.79,
"eval_steps_per_second": 10.236,
"step": 15000
},
{
"epoch": 0.65,
"learning_rate": 4.4425e-07,
"loss": 1.0987,
"step": 16000
},
{
"epoch": 0.65,
"eval_accuracy": 0.38962075848303396,
"eval_loss": 1.0837221145629883,
"eval_runtime": 61.1947,
"eval_samples_per_second": 81.87,
"eval_steps_per_second": 10.246,
"step": 16000
},
{
"epoch": 0.69,
"learning_rate": 4.7202777777777785e-07,
"loss": 1.0943,
"step": 17000
},
{
"epoch": 0.69,
"eval_accuracy": 0.4033932135728543,
"eval_loss": 1.081850290298462,
"eval_runtime": 61.2293,
"eval_samples_per_second": 81.824,
"eval_steps_per_second": 10.24,
"step": 17000
},
{
"epoch": 0.73,
"learning_rate": 4.997777777777779e-07,
"loss": 1.0966,
"step": 18000
},
{
"epoch": 0.73,
"eval_accuracy": 0.4031936127744511,
"eval_loss": 1.0839308500289917,
"eval_runtime": 61.2192,
"eval_samples_per_second": 81.837,
"eval_steps_per_second": 10.242,
"step": 18000
},
{
"epoch": 0.77,
"learning_rate": 5.275555555555556e-07,
"loss": 1.0935,
"step": 19000
},
{
"epoch": 0.77,
"eval_accuracy": 0.40099800399201596,
"eval_loss": 1.0814112424850464,
"eval_runtime": 61.1921,
"eval_samples_per_second": 81.873,
"eval_steps_per_second": 10.246,
"step": 19000
},
{
"epoch": 0.81,
"learning_rate": 5.553055555555556e-07,
"loss": 1.0939,
"step": 20000
},
{
"epoch": 0.81,
"eval_accuracy": 0.39241516966067863,
"eval_loss": 1.0840386152267456,
"eval_runtime": 61.2683,
"eval_samples_per_second": 81.772,
"eval_steps_per_second": 10.234,
"step": 20000
},
{
"epoch": 0.86,
"learning_rate": 5.830833333333334e-07,
"loss": 1.0917,
"step": 21000
},
{
"epoch": 0.86,
"eval_accuracy": 0.38303393213572856,
"eval_loss": 1.083555817604065,
"eval_runtime": 61.2974,
"eval_samples_per_second": 81.733,
"eval_steps_per_second": 10.229,
"step": 21000
},
{
"epoch": 0.9,
"learning_rate": 6.108333333333333e-07,
"loss": 1.0926,
"step": 22000
},
{
"epoch": 0.9,
"eval_accuracy": 0.39481037924151696,
"eval_loss": 1.080513834953308,
"eval_runtime": 61.4078,
"eval_samples_per_second": 81.586,
"eval_steps_per_second": 10.21,
"step": 22000
},
{
"epoch": 0.94,
"learning_rate": 6.386111111111112e-07,
"loss": 1.09,
"step": 23000
},
{
"epoch": 0.94,
"eval_accuracy": 0.4115768463073852,
"eval_loss": 1.076084017753601,
"eval_runtime": 61.6647,
"eval_samples_per_second": 81.246,
"eval_steps_per_second": 10.168,
"step": 23000
},
{
"epoch": 0.98,
"learning_rate": 6.663888888888889e-07,
"loss": 1.0879,
"step": 24000
},
{
"epoch": 0.98,
"eval_accuracy": 0.42734530938123755,
"eval_loss": 1.0709619522094727,
"eval_runtime": 61.3291,
"eval_samples_per_second": 81.69,
"eval_steps_per_second": 10.224,
"step": 24000
},
{
"epoch": 1.02,
"learning_rate": 6.94138888888889e-07,
"loss": 1.0846,
"step": 25000
},
{
"epoch": 1.02,
"eval_accuracy": 0.417564870259481,
"eval_loss": 1.0698238611221313,
"eval_runtime": 61.3091,
"eval_samples_per_second": 81.717,
"eval_steps_per_second": 10.227,
"step": 25000
},
{
"epoch": 1.06,
"learning_rate": 7.219166666666666e-07,
"loss": 1.0821,
"step": 26000
},
{
"epoch": 1.06,
"eval_accuracy": 0.43313373253493015,
"eval_loss": 1.065280795097351,
"eval_runtime": 61.4486,
"eval_samples_per_second": 81.532,
"eval_steps_per_second": 10.204,
"step": 26000
},
{
"epoch": 1.1,
"learning_rate": 7.496666666666667e-07,
"loss": 1.078,
"step": 27000
},
{
"epoch": 1.1,
"eval_accuracy": 0.4367265469061876,
"eval_loss": 1.059377908706665,
"eval_runtime": 61.6257,
"eval_samples_per_second": 81.297,
"eval_steps_per_second": 10.174,
"step": 27000
},
{
"epoch": 1.14,
"learning_rate": 7.774444444444445e-07,
"loss": 1.0707,
"step": 28000
},
{
"epoch": 1.14,
"eval_accuracy": 0.437125748502994,
"eval_loss": 1.0536556243896484,
"eval_runtime": 61.8255,
"eval_samples_per_second": 81.035,
"eval_steps_per_second": 10.141,
"step": 28000
},
{
"epoch": 1.18,
"learning_rate": 8.051944444444445e-07,
"loss": 1.0658,
"step": 29000
},
{
"epoch": 1.18,
"eval_accuracy": 0.4425149700598802,
"eval_loss": 1.0465552806854248,
"eval_runtime": 61.6813,
"eval_samples_per_second": 81.224,
"eval_steps_per_second": 10.165,
"step": 29000
},
{
"epoch": 1.22,
"learning_rate": 8.329722222222223e-07,
"loss": 1.0596,
"step": 30000
},
{
"epoch": 1.22,
"eval_accuracy": 0.44011976047904194,
"eval_loss": 1.042420506477356,
"eval_runtime": 61.7796,
"eval_samples_per_second": 81.095,
"eval_steps_per_second": 10.149,
"step": 30000
},
{
"epoch": 1.26,
"learning_rate": 8.607222222222223e-07,
"loss": 1.0665,
"step": 31000
},
{
"epoch": 1.26,
"eval_accuracy": 0.44011976047904194,
"eval_loss": 1.0469056367874146,
"eval_runtime": 62.7365,
"eval_samples_per_second": 79.858,
"eval_steps_per_second": 9.994,
"step": 31000
},
{
"epoch": 1.3,
"learning_rate": 8.885e-07,
"loss": 1.0586,
"step": 32000
},
{
"epoch": 1.3,
"eval_accuracy": 0.4363273453093812,
"eval_loss": 1.041648507118225,
"eval_runtime": 61.7699,
"eval_samples_per_second": 81.107,
"eval_steps_per_second": 10.151,
"step": 32000
},
{
"epoch": 1.34,
"learning_rate": 9.162500000000001e-07,
"loss": 1.0579,
"step": 33000
},
{
"epoch": 1.34,
"eval_accuracy": 0.44550898203592815,
"eval_loss": 1.0353455543518066,
"eval_runtime": 61.7392,
"eval_samples_per_second": 81.148,
"eval_steps_per_second": 10.156,
"step": 33000
},
{
"epoch": 1.39,
"learning_rate": 9.440277777777779e-07,
"loss": 1.0499,
"step": 34000
},
{
"epoch": 1.39,
"eval_accuracy": 0.45109780439121755,
"eval_loss": 1.0313138961791992,
"eval_runtime": 61.661,
"eval_samples_per_second": 81.251,
"eval_steps_per_second": 10.169,
"step": 34000
},
{
"epoch": 1.43,
"learning_rate": 9.718055555555557e-07,
"loss": 1.0551,
"step": 35000
},
{
"epoch": 1.43,
"eval_accuracy": 0.4481037924151697,
"eval_loss": 1.0282059907913208,
"eval_runtime": 61.7111,
"eval_samples_per_second": 81.185,
"eval_steps_per_second": 10.16,
"step": 35000
},
{
"epoch": 1.47,
"learning_rate": 9.995555555555557e-07,
"loss": 1.0509,
"step": 36000
},
{
"epoch": 1.47,
"eval_accuracy": 0.4568862275449102,
"eval_loss": 1.0246349573135376,
"eval_runtime": 61.4536,
"eval_samples_per_second": 81.525,
"eval_steps_per_second": 10.203,
"step": 36000
},
{
"epoch": 1.51,
"learning_rate": 1.0273333333333335e-06,
"loss": 1.0486,
"step": 37000
},
{
"epoch": 1.51,
"eval_accuracy": 0.45868263473053894,
"eval_loss": 1.0200064182281494,
"eval_runtime": 61.3644,
"eval_samples_per_second": 81.643,
"eval_steps_per_second": 10.218,
"step": 37000
},
{
"epoch": 1.55,
"learning_rate": 1.0550833333333334e-06,
"loss": 1.0468,
"step": 38000
},
{
"epoch": 1.55,
"eval_accuracy": 0.4564870259481038,
"eval_loss": 1.0205750465393066,
"eval_runtime": 61.6145,
"eval_samples_per_second": 81.312,
"eval_steps_per_second": 10.176,
"step": 38000
},
{
"epoch": 1.59,
"learning_rate": 1.0828611111111111e-06,
"loss": 1.0464,
"step": 39000
},
{
"epoch": 1.59,
"eval_accuracy": 0.4554890219560878,
"eval_loss": 1.021471619606018,
"eval_runtime": 61.6216,
"eval_samples_per_second": 81.303,
"eval_steps_per_second": 10.175,
"step": 39000
},
{
"epoch": 1.63,
"learning_rate": 1.1106111111111112e-06,
"loss": 1.0374,
"step": 40000
},
{
"epoch": 1.63,
"eval_accuracy": 0.45728542914171655,
"eval_loss": 1.0143259763717651,
"eval_runtime": 61.6469,
"eval_samples_per_second": 81.269,
"eval_steps_per_second": 10.171,
"step": 40000
},
{
"epoch": 1.67,
"learning_rate": 1.138388888888889e-06,
"loss": 1.0427,
"step": 41000
},
{
"epoch": 1.67,
"eval_accuracy": 0.45848303393213574,
"eval_loss": 1.0229008197784424,
"eval_runtime": 61.4018,
"eval_samples_per_second": 81.594,
"eval_steps_per_second": 10.211,
"step": 41000
},
{
"epoch": 1.71,
"learning_rate": 1.166138888888889e-06,
"loss": 1.0359,
"step": 42000
},
{
"epoch": 1.71,
"eval_accuracy": 0.4674650698602794,
"eval_loss": 1.0088660717010498,
"eval_runtime": 61.4607,
"eval_samples_per_second": 81.515,
"eval_steps_per_second": 10.202,
"step": 42000
},
{
"epoch": 1.75,
"learning_rate": 1.1939166666666668e-06,
"loss": 1.0382,
"step": 43000
},
{
"epoch": 1.75,
"eval_accuracy": 0.4489021956087824,
"eval_loss": 1.0299837589263916,
"eval_runtime": 61.6434,
"eval_samples_per_second": 81.274,
"eval_steps_per_second": 10.171,
"step": 43000
},
{
"epoch": 1.79,
"learning_rate": 1.2216944444444446e-06,
"loss": 1.0381,
"step": 44000
},
{
"epoch": 1.79,
"eval_accuracy": 0.4786427145708583,
"eval_loss": 1.0064666271209717,
"eval_runtime": 61.6779,
"eval_samples_per_second": 81.228,
"eval_steps_per_second": 10.166,
"step": 44000
},
{
"epoch": 1.83,
"learning_rate": 1.2494444444444447e-06,
"loss": 1.0369,
"step": 45000
},
{
"epoch": 1.83,
"eval_accuracy": 0.4716566866267465,
"eval_loss": 1.0102956295013428,
"eval_runtime": 61.6864,
"eval_samples_per_second": 81.217,
"eval_steps_per_second": 10.164,
"step": 45000
},
{
"epoch": 1.87,
"learning_rate": 1.2772222222222224e-06,
"loss": 1.0359,
"step": 46000
},
{
"epoch": 1.87,
"eval_accuracy": 0.4686626746506986,
"eval_loss": 1.011850118637085,
"eval_runtime": 61.3581,
"eval_samples_per_second": 81.652,
"eval_steps_per_second": 10.219,
"step": 46000
},
{
"epoch": 1.91,
"learning_rate": 1.3049722222222223e-06,
"loss": 1.0305,
"step": 47000
},
{
"epoch": 1.91,
"eval_accuracy": 0.48283433133732534,
"eval_loss": 1.0011930465698242,
"eval_runtime": 61.3161,
"eval_samples_per_second": 81.708,
"eval_steps_per_second": 10.226,
"step": 47000
},
{
"epoch": 1.96,
"learning_rate": 1.33275e-06,
"loss": 1.035,
"step": 48000
},
{
"epoch": 1.96,
"eval_accuracy": 0.4467065868263473,
"eval_loss": 1.0333038568496704,
"eval_runtime": 61.3247,
"eval_samples_per_second": 81.696,
"eval_steps_per_second": 10.224,
"step": 48000
},
{
"epoch": 2.0,
"learning_rate": 1.3605277777777778e-06,
"loss": 1.027,
"step": 49000
},
{
"epoch": 2.0,
"eval_accuracy": 0.48862275449101794,
"eval_loss": 0.9966129064559937,
"eval_runtime": 61.479,
"eval_samples_per_second": 81.491,
"eval_steps_per_second": 10.199,
"step": 49000
},
{
"epoch": 2.04,
"learning_rate": 1.388277777777778e-06,
"loss": 1.0271,
"step": 50000
},
{
"epoch": 2.04,
"eval_accuracy": 0.4744510978043912,
"eval_loss": 1.006515622138977,
"eval_runtime": 61.3701,
"eval_samples_per_second": 81.636,
"eval_steps_per_second": 10.217,
"step": 50000
},
{
"epoch": 2.08,
"learning_rate": 1.4160555555555555e-06,
"loss": 1.0271,
"step": 51000
},
{
"epoch": 2.08,
"eval_accuracy": 0.49141716566866267,
"eval_loss": 0.9953572154045105,
"eval_runtime": 61.4077,
"eval_samples_per_second": 81.586,
"eval_steps_per_second": 10.21,
"step": 51000
},
{
"epoch": 2.12,
"learning_rate": 1.4438055555555558e-06,
"loss": 1.0233,
"step": 52000
},
{
"epoch": 2.12,
"eval_accuracy": 0.4728542914171657,
"eval_loss": 1.017760992050171,
"eval_runtime": 61.6076,
"eval_samples_per_second": 81.321,
"eval_steps_per_second": 10.177,
"step": 52000
},
{
"epoch": 2.16,
"learning_rate": 1.4715833333333333e-06,
"loss": 1.0275,
"step": 53000
},
{
"epoch": 2.16,
"eval_accuracy": 0.49860279441117766,
"eval_loss": 0.9917733073234558,
"eval_runtime": 61.3839,
"eval_samples_per_second": 81.617,
"eval_steps_per_second": 10.214,
"step": 53000
}
],
"logging_steps": 1000,
"max_steps": 10000000,
"num_train_epochs": 408,
"save_steps": 1000,
"total_flos": 1.107882993207214e+17,
"trial_name": null,
"trial_params": null
}