tmnam20's picture
Training in progress, step 54000, checkpoint
3105033
raw
history blame
20.5 kB
{
"best_metric": 1.0106219053268433,
"best_model_checkpoint": "/kaggle/output/checkpoint-54000",
"epoch": 2.2001303780964796,
"eval_steps": 1000,
"global_step": 54000,
"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": 1.1078,
"step": 1
},
{
"epoch": 0.04,
"learning_rate": 2.7750000000000004e-08,
"loss": 1.128,
"step": 1000
},
{
"epoch": 0.04,
"eval_accuracy": 0.330938123752495,
"eval_loss": 1.1083118915557861,
"eval_runtime": 12.3979,
"eval_samples_per_second": 404.1,
"eval_steps_per_second": 50.573,
"step": 1000
},
{
"epoch": 0.08,
"learning_rate": 5.5527777777777784e-08,
"loss": 1.1164,
"step": 2000
},
{
"epoch": 0.08,
"eval_accuracy": 0.3323353293413174,
"eval_loss": 1.102672815322876,
"eval_runtime": 12.2861,
"eval_samples_per_second": 407.778,
"eval_steps_per_second": 51.033,
"step": 2000
},
{
"epoch": 0.12,
"learning_rate": 8.327777777777778e-08,
"loss": 1.1158,
"step": 3000
},
{
"epoch": 0.12,
"eval_accuracy": 0.34211576846307384,
"eval_loss": 1.1024774312973022,
"eval_runtime": 12.3746,
"eval_samples_per_second": 404.863,
"eval_steps_per_second": 50.668,
"step": 3000
},
{
"epoch": 0.16,
"learning_rate": 1.1105555555555557e-07,
"loss": 1.1115,
"step": 4000
},
{
"epoch": 0.16,
"eval_accuracy": 0.34890219560878244,
"eval_loss": 1.1012382507324219,
"eval_runtime": 12.5405,
"eval_samples_per_second": 399.505,
"eval_steps_per_second": 49.998,
"step": 4000
},
{
"epoch": 0.2,
"learning_rate": 1.3880555555555558e-07,
"loss": 1.1141,
"step": 5000
},
{
"epoch": 0.2,
"eval_accuracy": 0.3534930139720559,
"eval_loss": 1.1002192497253418,
"eval_runtime": 12.3001,
"eval_samples_per_second": 407.314,
"eval_steps_per_second": 50.975,
"step": 5000
},
{
"epoch": 0.24,
"learning_rate": 1.6658333333333335e-07,
"loss": 1.112,
"step": 6000
},
{
"epoch": 0.24,
"eval_accuracy": 0.3524950099800399,
"eval_loss": 1.100056767463684,
"eval_runtime": 12.4062,
"eval_samples_per_second": 403.831,
"eval_steps_per_second": 50.539,
"step": 6000
},
{
"epoch": 0.29,
"learning_rate": 1.9433333333333334e-07,
"loss": 1.1115,
"step": 7000
},
{
"epoch": 0.29,
"eval_accuracy": 0.34331337325349304,
"eval_loss": 1.098986029624939,
"eval_runtime": 12.4456,
"eval_samples_per_second": 402.551,
"eval_steps_per_second": 50.379,
"step": 7000
},
{
"epoch": 0.33,
"learning_rate": 2.2211111111111114e-07,
"loss": 1.109,
"step": 8000
},
{
"epoch": 0.33,
"eval_accuracy": 0.3347305389221557,
"eval_loss": 1.098678469657898,
"eval_runtime": 12.5978,
"eval_samples_per_second": 397.688,
"eval_steps_per_second": 49.771,
"step": 8000
},
{
"epoch": 0.37,
"learning_rate": 2.4986111111111113e-07,
"loss": 1.1117,
"step": 9000
},
{
"epoch": 0.37,
"eval_accuracy": 0.34311377245508984,
"eval_loss": 1.0984419584274292,
"eval_runtime": 12.313,
"eval_samples_per_second": 406.886,
"eval_steps_per_second": 50.922,
"step": 9000
},
{
"epoch": 0.41,
"learning_rate": 2.776388888888889e-07,
"loss": 1.1081,
"step": 10000
},
{
"epoch": 0.41,
"eval_accuracy": 0.35429141716566864,
"eval_loss": 1.0968743562698364,
"eval_runtime": 12.4088,
"eval_samples_per_second": 403.744,
"eval_steps_per_second": 50.528,
"step": 10000
},
{
"epoch": 0.45,
"learning_rate": 3.053888888888889e-07,
"loss": 1.107,
"step": 11000
},
{
"epoch": 0.45,
"eval_accuracy": 0.3722554890219561,
"eval_loss": 1.0950665473937988,
"eval_runtime": 12.5775,
"eval_samples_per_second": 398.33,
"eval_steps_per_second": 49.851,
"step": 11000
},
{
"epoch": 0.49,
"learning_rate": 3.331666666666667e-07,
"loss": 1.1057,
"step": 12000
},
{
"epoch": 0.49,
"eval_accuracy": 0.37544910179640717,
"eval_loss": 1.0940738916397095,
"eval_runtime": 12.5186,
"eval_samples_per_second": 400.203,
"eval_steps_per_second": 50.085,
"step": 12000
},
{
"epoch": 0.53,
"learning_rate": 3.609166666666667e-07,
"loss": 1.1069,
"step": 13000
},
{
"epoch": 0.53,
"eval_accuracy": 0.3401197604790419,
"eval_loss": 1.0937162637710571,
"eval_runtime": 12.2594,
"eval_samples_per_second": 408.666,
"eval_steps_per_second": 51.144,
"step": 13000
},
{
"epoch": 0.57,
"learning_rate": 3.886944444444445e-07,
"loss": 1.1046,
"step": 14000
},
{
"epoch": 0.57,
"eval_accuracy": 0.37604790419161677,
"eval_loss": 1.0926916599273682,
"eval_runtime": 12.256,
"eval_samples_per_second": 408.78,
"eval_steps_per_second": 51.159,
"step": 14000
},
{
"epoch": 0.61,
"learning_rate": 4.164444444444445e-07,
"loss": 1.1046,
"step": 15000
},
{
"epoch": 0.61,
"eval_accuracy": 0.39201596806387223,
"eval_loss": 1.0912976264953613,
"eval_runtime": 12.5019,
"eval_samples_per_second": 400.74,
"eval_steps_per_second": 50.152,
"step": 15000
},
{
"epoch": 0.65,
"learning_rate": 4.442222222222223e-07,
"loss": 1.1052,
"step": 16000
},
{
"epoch": 0.65,
"eval_accuracy": 0.393812375249501,
"eval_loss": 1.0901482105255127,
"eval_runtime": 12.3407,
"eval_samples_per_second": 405.973,
"eval_steps_per_second": 50.807,
"step": 16000
},
{
"epoch": 0.69,
"learning_rate": 4.7197222222222224e-07,
"loss": 1.1011,
"step": 17000
},
{
"epoch": 0.69,
"eval_accuracy": 0.3932135728542914,
"eval_loss": 1.0901678800582886,
"eval_runtime": 12.3644,
"eval_samples_per_second": 405.197,
"eval_steps_per_second": 50.71,
"step": 17000
},
{
"epoch": 0.73,
"learning_rate": 4.997500000000001e-07,
"loss": 1.1,
"step": 18000
},
{
"epoch": 0.73,
"eval_accuracy": 0.38143712574850297,
"eval_loss": 1.0897998809814453,
"eval_runtime": 12.3516,
"eval_samples_per_second": 405.615,
"eval_steps_per_second": 50.763,
"step": 18000
},
{
"epoch": 0.77,
"learning_rate": 5.275277777777778e-07,
"loss": 1.1007,
"step": 19000
},
{
"epoch": 0.77,
"eval_accuracy": 0.40678642714570856,
"eval_loss": 1.0876121520996094,
"eval_runtime": 12.6608,
"eval_samples_per_second": 395.708,
"eval_steps_per_second": 49.523,
"step": 19000
},
{
"epoch": 0.81,
"learning_rate": 5.552777777777778e-07,
"loss": 1.1013,
"step": 20000
},
{
"epoch": 0.81,
"eval_accuracy": 0.4053892215568862,
"eval_loss": 1.0876350402832031,
"eval_runtime": 12.3243,
"eval_samples_per_second": 406.513,
"eval_steps_per_second": 50.875,
"step": 20000
},
{
"epoch": 0.86,
"learning_rate": 5.830555555555556e-07,
"loss": 1.0987,
"step": 21000
},
{
"epoch": 0.86,
"eval_accuracy": 0.40858283433133735,
"eval_loss": 1.085940957069397,
"eval_runtime": 12.3217,
"eval_samples_per_second": 406.598,
"eval_steps_per_second": 50.886,
"step": 21000
},
{
"epoch": 0.9,
"learning_rate": 6.108055555555556e-07,
"loss": 1.0986,
"step": 22000
},
{
"epoch": 0.9,
"eval_accuracy": 0.3972055888223553,
"eval_loss": 1.085978388786316,
"eval_runtime": 12.4285,
"eval_samples_per_second": 403.107,
"eval_steps_per_second": 50.449,
"step": 22000
},
{
"epoch": 0.94,
"learning_rate": 6.385833333333334e-07,
"loss": 1.0944,
"step": 23000
},
{
"epoch": 0.94,
"eval_accuracy": 0.41097804391217563,
"eval_loss": 1.0837496519088745,
"eval_runtime": 12.7175,
"eval_samples_per_second": 393.946,
"eval_steps_per_second": 49.302,
"step": 23000
},
{
"epoch": 0.98,
"learning_rate": 6.663611111111112e-07,
"loss": 1.095,
"step": 24000
},
{
"epoch": 0.98,
"eval_accuracy": 0.4121756487025948,
"eval_loss": 1.0820704698562622,
"eval_runtime": 12.3323,
"eval_samples_per_second": 406.251,
"eval_steps_per_second": 50.842,
"step": 24000
},
{
"epoch": 1.02,
"learning_rate": 6.941111111111112e-07,
"loss": 1.0915,
"step": 25000
},
{
"epoch": 1.02,
"eval_accuracy": 0.4177644710578842,
"eval_loss": 1.0762931108474731,
"eval_runtime": 12.3887,
"eval_samples_per_second": 404.399,
"eval_steps_per_second": 50.61,
"step": 25000
},
{
"epoch": 1.06,
"learning_rate": 7.218888888888889e-07,
"loss": 1.0889,
"step": 26000
},
{
"epoch": 1.06,
"eval_accuracy": 0.43253493013972055,
"eval_loss": 1.0695326328277588,
"eval_runtime": 12.5611,
"eval_samples_per_second": 398.85,
"eval_steps_per_second": 49.916,
"step": 26000
},
{
"epoch": 1.1,
"learning_rate": 7.49638888888889e-07,
"loss": 1.0827,
"step": 27000
},
{
"epoch": 1.1,
"eval_accuracy": 0.4343313373253493,
"eval_loss": 1.0662548542022705,
"eval_runtime": 12.641,
"eval_samples_per_second": 396.328,
"eval_steps_per_second": 49.6,
"step": 27000
},
{
"epoch": 1.14,
"learning_rate": 7.774166666666668e-07,
"loss": 1.0765,
"step": 28000
},
{
"epoch": 1.14,
"eval_accuracy": 0.4315369261477046,
"eval_loss": 1.0609794855117798,
"eval_runtime": 12.4637,
"eval_samples_per_second": 401.967,
"eval_steps_per_second": 50.306,
"step": 28000
},
{
"epoch": 1.18,
"learning_rate": 8.051666666666667e-07,
"loss": 1.0736,
"step": 29000
},
{
"epoch": 1.18,
"eval_accuracy": 0.4377245508982036,
"eval_loss": 1.0587859153747559,
"eval_runtime": 12.3801,
"eval_samples_per_second": 404.683,
"eval_steps_per_second": 50.646,
"step": 29000
},
{
"epoch": 1.22,
"learning_rate": 8.329444444444445e-07,
"loss": 1.0726,
"step": 30000
},
{
"epoch": 1.22,
"eval_accuracy": 0.4249500998003992,
"eval_loss": 1.0559338331222534,
"eval_runtime": 12.5945,
"eval_samples_per_second": 397.792,
"eval_steps_per_second": 49.784,
"step": 30000
},
{
"epoch": 1.26,
"learning_rate": 8.606944444444445e-07,
"loss": 1.0751,
"step": 31000
},
{
"epoch": 1.26,
"eval_accuracy": 0.4427145708582834,
"eval_loss": 1.0576916933059692,
"eval_runtime": 12.6324,
"eval_samples_per_second": 396.599,
"eval_steps_per_second": 49.634,
"step": 31000
},
{
"epoch": 1.3,
"learning_rate": 8.884722222222224e-07,
"loss": 1.0708,
"step": 32000
},
{
"epoch": 1.3,
"eval_accuracy": 0.42634730538922155,
"eval_loss": 1.0571542978286743,
"eval_runtime": 12.4696,
"eval_samples_per_second": 401.776,
"eval_steps_per_second": 50.282,
"step": 32000
},
{
"epoch": 1.34,
"learning_rate": 9.162222222222223e-07,
"loss": 1.0711,
"step": 33000
},
{
"epoch": 1.34,
"eval_accuracy": 0.4419161676646707,
"eval_loss": 1.0544849634170532,
"eval_runtime": 12.415,
"eval_samples_per_second": 403.544,
"eval_steps_per_second": 50.503,
"step": 33000
},
{
"epoch": 1.39,
"learning_rate": 9.440000000000001e-07,
"loss": 1.0666,
"step": 34000
},
{
"epoch": 1.39,
"eval_accuracy": 0.43333333333333335,
"eval_loss": 1.05018150806427,
"eval_runtime": 12.6226,
"eval_samples_per_second": 396.906,
"eval_steps_per_second": 49.673,
"step": 34000
},
{
"epoch": 1.43,
"learning_rate": 9.7175e-07,
"loss": 1.0698,
"step": 35000
},
{
"epoch": 1.43,
"eval_accuracy": 0.4377245508982036,
"eval_loss": 1.0488406419754028,
"eval_runtime": 12.6246,
"eval_samples_per_second": 396.844,
"eval_steps_per_second": 49.665,
"step": 35000
},
{
"epoch": 1.47,
"learning_rate": 9.995277777777778e-07,
"loss": 1.067,
"step": 36000
},
{
"epoch": 1.47,
"eval_accuracy": 0.4407185628742515,
"eval_loss": 1.047359824180603,
"eval_runtime": 12.4039,
"eval_samples_per_second": 403.906,
"eval_steps_per_second": 50.549,
"step": 36000
},
{
"epoch": 1.51,
"learning_rate": 1.027277777777778e-06,
"loss": 1.0636,
"step": 37000
},
{
"epoch": 1.51,
"eval_accuracy": 0.44451097804391215,
"eval_loss": 1.0455204248428345,
"eval_runtime": 12.3836,
"eval_samples_per_second": 404.569,
"eval_steps_per_second": 50.632,
"step": 37000
},
{
"epoch": 1.55,
"learning_rate": 1.0550555555555557e-06,
"loss": 1.0658,
"step": 38000
},
{
"epoch": 1.55,
"eval_accuracy": 0.4407185628742515,
"eval_loss": 1.0461794137954712,
"eval_runtime": 12.5546,
"eval_samples_per_second": 399.057,
"eval_steps_per_second": 49.942,
"step": 38000
},
{
"epoch": 1.59,
"learning_rate": 1.0828333333333334e-06,
"loss": 1.0635,
"step": 39000
},
{
"epoch": 1.59,
"eval_accuracy": 0.4367265469061876,
"eval_loss": 1.0461270809173584,
"eval_runtime": 12.5509,
"eval_samples_per_second": 399.173,
"eval_steps_per_second": 49.956,
"step": 39000
},
{
"epoch": 1.63,
"learning_rate": 1.1105833333333335e-06,
"loss": 1.0596,
"step": 40000
},
{
"epoch": 1.63,
"eval_accuracy": 0.449500998003992,
"eval_loss": 1.0418727397918701,
"eval_runtime": 12.4044,
"eval_samples_per_second": 403.889,
"eval_steps_per_second": 50.547,
"step": 40000
},
{
"epoch": 1.67,
"learning_rate": 1.1383611111111113e-06,
"loss": 1.0595,
"step": 41000
},
{
"epoch": 1.67,
"eval_accuracy": 0.4409181636726547,
"eval_loss": 1.0417557954788208,
"eval_runtime": 12.3642,
"eval_samples_per_second": 405.202,
"eval_steps_per_second": 50.711,
"step": 41000
},
{
"epoch": 1.71,
"learning_rate": 1.1661111111111111e-06,
"loss": 1.054,
"step": 42000
},
{
"epoch": 1.71,
"eval_accuracy": 0.4357285429141717,
"eval_loss": 1.0383535623550415,
"eval_runtime": 12.3634,
"eval_samples_per_second": 405.23,
"eval_steps_per_second": 50.714,
"step": 42000
},
{
"epoch": 1.75,
"learning_rate": 1.193888888888889e-06,
"loss": 1.0543,
"step": 43000
},
{
"epoch": 1.75,
"eval_accuracy": 0.43812375249500995,
"eval_loss": 1.0463169813156128,
"eval_runtime": 12.6398,
"eval_samples_per_second": 396.366,
"eval_steps_per_second": 49.605,
"step": 43000
},
{
"epoch": 1.79,
"learning_rate": 1.221638888888889e-06,
"loss": 1.0579,
"step": 44000
},
{
"epoch": 1.79,
"eval_accuracy": 0.4415169660678643,
"eval_loss": 1.0388957262039185,
"eval_runtime": 12.4542,
"eval_samples_per_second": 402.274,
"eval_steps_per_second": 50.344,
"step": 44000
},
{
"epoch": 1.83,
"learning_rate": 1.2494166666666668e-06,
"loss": 1.0561,
"step": 45000
},
{
"epoch": 1.83,
"eval_accuracy": 0.46187624750499,
"eval_loss": 1.031552791595459,
"eval_runtime": 12.4082,
"eval_samples_per_second": 403.764,
"eval_steps_per_second": 50.531,
"step": 45000
},
{
"epoch": 1.87,
"learning_rate": 1.2771666666666668e-06,
"loss": 1.0516,
"step": 46000
},
{
"epoch": 1.87,
"eval_accuracy": 0.4538922155688623,
"eval_loss": 1.0312999486923218,
"eval_runtime": 12.3695,
"eval_samples_per_second": 405.029,
"eval_steps_per_second": 50.689,
"step": 46000
},
{
"epoch": 1.91,
"learning_rate": 1.3049444444444446e-06,
"loss": 1.0471,
"step": 47000
},
{
"epoch": 1.91,
"eval_accuracy": 0.46387225548902195,
"eval_loss": 1.027535319328308,
"eval_runtime": 12.5846,
"eval_samples_per_second": 398.107,
"eval_steps_per_second": 49.823,
"step": 47000
},
{
"epoch": 1.96,
"learning_rate": 1.3326944444444447e-06,
"loss": 1.054,
"step": 48000
},
{
"epoch": 1.96,
"eval_accuracy": 0.4568862275449102,
"eval_loss": 1.0308438539505005,
"eval_runtime": 12.5646,
"eval_samples_per_second": 398.74,
"eval_steps_per_second": 49.902,
"step": 48000
},
{
"epoch": 2.0,
"learning_rate": 1.3604722222222224e-06,
"loss": 1.0445,
"step": 49000
},
{
"epoch": 2.0,
"eval_accuracy": 0.47045908183632734,
"eval_loss": 1.0230755805969238,
"eval_runtime": 12.4305,
"eval_samples_per_second": 403.041,
"eval_steps_per_second": 50.44,
"step": 49000
},
{
"epoch": 2.04,
"learning_rate": 1.3882222222222223e-06,
"loss": 1.0479,
"step": 50000
},
{
"epoch": 2.04,
"eval_accuracy": 0.47305389221556887,
"eval_loss": 1.0201445817947388,
"eval_runtime": 12.3672,
"eval_samples_per_second": 405.103,
"eval_steps_per_second": 50.698,
"step": 50000
},
{
"epoch": 2.08,
"learning_rate": 1.416e-06,
"loss": 1.0482,
"step": 51000
},
{
"epoch": 2.08,
"eval_accuracy": 0.4678642714570858,
"eval_loss": 1.0167714357376099,
"eval_runtime": 12.5982,
"eval_samples_per_second": 397.677,
"eval_steps_per_second": 49.769,
"step": 51000
},
{
"epoch": 2.12,
"learning_rate": 1.4437500000000002e-06,
"loss": 1.0406,
"step": 52000
},
{
"epoch": 2.12,
"eval_accuracy": 0.4722554890219561,
"eval_loss": 1.0175116062164307,
"eval_runtime": 12.6041,
"eval_samples_per_second": 397.49,
"eval_steps_per_second": 49.746,
"step": 52000
},
{
"epoch": 2.16,
"learning_rate": 1.471527777777778e-06,
"loss": 1.0457,
"step": 53000
},
{
"epoch": 2.16,
"eval_accuracy": 0.4748502994011976,
"eval_loss": 1.010857343673706,
"eval_runtime": 12.3728,
"eval_samples_per_second": 404.921,
"eval_steps_per_second": 50.676,
"step": 53000
},
{
"epoch": 2.2,
"learning_rate": 1.4993055555555557e-06,
"loss": 1.0399,
"step": 54000
},
{
"epoch": 2.2,
"eval_accuracy": 0.48522954091816367,
"eval_loss": 1.0106219053268433,
"eval_runtime": 12.3952,
"eval_samples_per_second": 404.188,
"eval_steps_per_second": 50.584,
"step": 54000
}
],
"logging_steps": 1000,
"max_steps": 10000000,
"num_train_epochs": 408,
"save_steps": 1000,
"total_flos": 1.881565710306509e+16,
"trial_name": null,
"trial_params": null
}