tmnam20's picture
Training in progress, step 113000, checkpoint
8c93e51
{
"best_metric": 0.9604237079620361,
"best_model_checkpoint": "/kaggle/output/checkpoint-112000",
"epoch": 4.603976531942633,
"eval_steps": 1000,
"global_step": 113000,
"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": 1.2358,
"step": 1
},
{
"epoch": 0.04,
"learning_rate": 2.7750000000000004e-08,
"loss": 1.1519,
"step": 1000
},
{
"epoch": 0.04,
"eval_accuracy": 0.3201596806387226,
"eval_loss": 1.1120171546936035,
"eval_runtime": 29.027,
"eval_samples_per_second": 172.598,
"eval_steps_per_second": 21.601,
"step": 1000
},
{
"epoch": 0.08,
"learning_rate": 5.5527777777777784e-08,
"loss": 1.128,
"step": 2000
},
{
"epoch": 0.08,
"eval_accuracy": 0.31796407185628744,
"eval_loss": 1.108055830001831,
"eval_runtime": 28.9944,
"eval_samples_per_second": 172.792,
"eval_steps_per_second": 21.625,
"step": 2000
},
{
"epoch": 0.12,
"learning_rate": 8.327777777777778e-08,
"loss": 1.122,
"step": 3000
},
{
"epoch": 0.12,
"eval_accuracy": 0.3345309381237525,
"eval_loss": 1.1074285507202148,
"eval_runtime": 28.9901,
"eval_samples_per_second": 172.818,
"eval_steps_per_second": 21.628,
"step": 3000
},
{
"epoch": 0.16,
"learning_rate": 1.1105555555555557e-07,
"loss": 1.1196,
"step": 4000
},
{
"epoch": 0.16,
"eval_accuracy": 0.3337325349301397,
"eval_loss": 1.1045424938201904,
"eval_runtime": 28.9791,
"eval_samples_per_second": 172.883,
"eval_steps_per_second": 21.636,
"step": 4000
},
{
"epoch": 0.2,
"learning_rate": 1.3880555555555558e-07,
"loss": 1.1169,
"step": 5000
},
{
"epoch": 0.2,
"eval_accuracy": 0.3397205588822355,
"eval_loss": 1.1033179759979248,
"eval_runtime": 29.0295,
"eval_samples_per_second": 172.583,
"eval_steps_per_second": 21.599,
"step": 5000
},
{
"epoch": 0.24,
"learning_rate": 1.6658333333333335e-07,
"loss": 1.118,
"step": 6000
},
{
"epoch": 0.24,
"eval_accuracy": 0.3401197604790419,
"eval_loss": 1.1030455827713013,
"eval_runtime": 28.9748,
"eval_samples_per_second": 172.909,
"eval_steps_per_second": 21.639,
"step": 6000
},
{
"epoch": 0.29,
"learning_rate": 1.9433333333333334e-07,
"loss": 1.1135,
"step": 7000
},
{
"epoch": 0.29,
"eval_accuracy": 0.35568862275449104,
"eval_loss": 1.0998502969741821,
"eval_runtime": 28.9964,
"eval_samples_per_second": 172.78,
"eval_steps_per_second": 21.623,
"step": 7000
},
{
"epoch": 0.33,
"learning_rate": 2.2211111111111114e-07,
"loss": 1.1136,
"step": 8000
},
{
"epoch": 0.33,
"eval_accuracy": 0.34510978043912177,
"eval_loss": 1.098555564880371,
"eval_runtime": 28.983,
"eval_samples_per_second": 172.86,
"eval_steps_per_second": 21.633,
"step": 8000
},
{
"epoch": 0.37,
"learning_rate": 2.4986111111111113e-07,
"loss": 1.11,
"step": 9000
},
{
"epoch": 0.37,
"eval_accuracy": 0.3411177644710579,
"eval_loss": 1.1035741567611694,
"eval_runtime": 28.9758,
"eval_samples_per_second": 172.903,
"eval_steps_per_second": 21.639,
"step": 9000
},
{
"epoch": 0.41,
"learning_rate": 2.776388888888889e-07,
"loss": 1.1076,
"step": 10000
},
{
"epoch": 0.41,
"eval_accuracy": 0.35788423153692617,
"eval_loss": 1.0970662832260132,
"eval_runtime": 29.0262,
"eval_samples_per_second": 172.603,
"eval_steps_per_second": 21.601,
"step": 10000
},
{
"epoch": 0.45,
"learning_rate": 3.0541666666666667e-07,
"loss": 1.1067,
"step": 11000
},
{
"epoch": 0.45,
"eval_accuracy": 0.37425149700598803,
"eval_loss": 1.094608187675476,
"eval_runtime": 28.8738,
"eval_samples_per_second": 173.513,
"eval_steps_per_second": 21.715,
"step": 11000
},
{
"epoch": 0.49,
"learning_rate": 3.3319444444444444e-07,
"loss": 1.1032,
"step": 12000
},
{
"epoch": 0.49,
"eval_accuracy": 0.37924151696606784,
"eval_loss": 1.092067837715149,
"eval_runtime": 28.9514,
"eval_samples_per_second": 173.049,
"eval_steps_per_second": 21.657,
"step": 12000
},
{
"epoch": 0.53,
"learning_rate": 3.6094444444444446e-07,
"loss": 1.1052,
"step": 13000
},
{
"epoch": 0.53,
"eval_accuracy": 0.3435129740518962,
"eval_loss": 1.0962508916854858,
"eval_runtime": 29.2203,
"eval_samples_per_second": 171.456,
"eval_steps_per_second": 21.458,
"step": 13000
},
{
"epoch": 0.57,
"learning_rate": 3.8872222222222223e-07,
"loss": 1.1015,
"step": 14000
},
{
"epoch": 0.57,
"eval_accuracy": 0.36806387225548903,
"eval_loss": 1.0941067934036255,
"eval_runtime": 29.2088,
"eval_samples_per_second": 171.524,
"eval_steps_per_second": 21.466,
"step": 14000
},
{
"epoch": 0.61,
"learning_rate": 4.1650000000000006e-07,
"loss": 1.0986,
"step": 15000
},
{
"epoch": 0.61,
"eval_accuracy": 0.3718562874251497,
"eval_loss": 1.0906143188476562,
"eval_runtime": 28.9463,
"eval_samples_per_second": 173.079,
"eval_steps_per_second": 21.661,
"step": 15000
},
{
"epoch": 0.65,
"learning_rate": 4.4425e-07,
"loss": 1.0984,
"step": 16000
},
{
"epoch": 0.65,
"eval_accuracy": 0.38183632734530937,
"eval_loss": 1.0862348079681396,
"eval_runtime": 28.9909,
"eval_samples_per_second": 172.813,
"eval_steps_per_second": 21.627,
"step": 16000
},
{
"epoch": 0.69,
"learning_rate": 4.7202777777777785e-07,
"loss": 1.0959,
"step": 17000
},
{
"epoch": 0.69,
"eval_accuracy": 0.40079840319361276,
"eval_loss": 1.0846956968307495,
"eval_runtime": 28.9606,
"eval_samples_per_second": 172.994,
"eval_steps_per_second": 21.65,
"step": 17000
},
{
"epoch": 0.73,
"learning_rate": 4.998055555555556e-07,
"loss": 1.0965,
"step": 18000
},
{
"epoch": 0.73,
"eval_accuracy": 0.4043912175648703,
"eval_loss": 1.086153507232666,
"eval_runtime": 28.8411,
"eval_samples_per_second": 173.71,
"eval_steps_per_second": 21.74,
"step": 18000
},
{
"epoch": 0.77,
"learning_rate": 5.275555555555556e-07,
"loss": 1.0966,
"step": 19000
},
{
"epoch": 0.77,
"eval_accuracy": 0.4041916167664671,
"eval_loss": 1.0824861526489258,
"eval_runtime": 28.9652,
"eval_samples_per_second": 172.966,
"eval_steps_per_second": 21.647,
"step": 19000
},
{
"epoch": 0.81,
"learning_rate": 5.553333333333334e-07,
"loss": 1.0943,
"step": 20000
},
{
"epoch": 0.81,
"eval_accuracy": 0.4103792415169661,
"eval_loss": 1.0817394256591797,
"eval_runtime": 28.9548,
"eval_samples_per_second": 173.028,
"eval_steps_per_second": 21.654,
"step": 20000
},
{
"epoch": 0.86,
"learning_rate": 5.830833333333334e-07,
"loss": 1.0898,
"step": 21000
},
{
"epoch": 0.86,
"eval_accuracy": 0.41357285429141716,
"eval_loss": 1.0764833688735962,
"eval_runtime": 28.9932,
"eval_samples_per_second": 172.799,
"eval_steps_per_second": 21.626,
"step": 21000
},
{
"epoch": 0.9,
"learning_rate": 6.108611111111111e-07,
"loss": 1.0877,
"step": 22000
},
{
"epoch": 0.9,
"eval_accuracy": 0.4293413173652695,
"eval_loss": 1.0701513290405273,
"eval_runtime": 28.9628,
"eval_samples_per_second": 172.98,
"eval_steps_per_second": 21.648,
"step": 22000
},
{
"epoch": 0.94,
"learning_rate": 6.386388888888889e-07,
"loss": 1.0826,
"step": 23000
},
{
"epoch": 0.94,
"eval_accuracy": 0.4285429141716567,
"eval_loss": 1.0636069774627686,
"eval_runtime": 28.9704,
"eval_samples_per_second": 172.935,
"eval_steps_per_second": 21.643,
"step": 23000
},
{
"epoch": 0.98,
"learning_rate": 6.664166666666667e-07,
"loss": 1.0781,
"step": 24000
},
{
"epoch": 0.98,
"eval_accuracy": 0.43213572854291415,
"eval_loss": 1.059762954711914,
"eval_runtime": 29.0265,
"eval_samples_per_second": 172.601,
"eval_steps_per_second": 21.601,
"step": 24000
},
{
"epoch": 1.02,
"learning_rate": 6.941666666666667e-07,
"loss": 1.0728,
"step": 25000
},
{
"epoch": 1.02,
"eval_accuracy": 0.4373253493013972,
"eval_loss": 1.056991696357727,
"eval_runtime": 28.9761,
"eval_samples_per_second": 172.901,
"eval_steps_per_second": 21.639,
"step": 25000
},
{
"epoch": 1.06,
"learning_rate": 7.219444444444444e-07,
"loss": 1.0719,
"step": 26000
},
{
"epoch": 1.06,
"eval_accuracy": 0.43193612774451096,
"eval_loss": 1.0573900938034058,
"eval_runtime": 28.9853,
"eval_samples_per_second": 172.846,
"eval_steps_per_second": 21.632,
"step": 26000
},
{
"epoch": 1.1,
"learning_rate": 7.496944444444444e-07,
"loss": 1.0683,
"step": 27000
},
{
"epoch": 1.1,
"eval_accuracy": 0.4305389221556886,
"eval_loss": 1.0567399263381958,
"eval_runtime": 28.9611,
"eval_samples_per_second": 172.991,
"eval_steps_per_second": 21.65,
"step": 27000
},
{
"epoch": 1.14,
"learning_rate": 7.774722222222223e-07,
"loss": 1.0669,
"step": 28000
},
{
"epoch": 1.14,
"eval_accuracy": 0.4297405189620758,
"eval_loss": 1.05367910861969,
"eval_runtime": 28.9913,
"eval_samples_per_second": 172.81,
"eval_steps_per_second": 21.627,
"step": 28000
},
{
"epoch": 1.18,
"learning_rate": 8.052222222222223e-07,
"loss": 1.0629,
"step": 29000
},
{
"epoch": 1.18,
"eval_accuracy": 0.4303393213572854,
"eval_loss": 1.0497413873672485,
"eval_runtime": 28.9605,
"eval_samples_per_second": 172.994,
"eval_steps_per_second": 21.65,
"step": 29000
},
{
"epoch": 1.22,
"learning_rate": 8.330000000000001e-07,
"loss": 1.0614,
"step": 30000
},
{
"epoch": 1.22,
"eval_accuracy": 0.43173652694610776,
"eval_loss": 1.046238899230957,
"eval_runtime": 28.9896,
"eval_samples_per_second": 172.821,
"eval_steps_per_second": 21.628,
"step": 30000
},
{
"epoch": 1.26,
"learning_rate": 8.607500000000001e-07,
"loss": 1.0669,
"step": 31000
},
{
"epoch": 1.26,
"eval_accuracy": 0.43233532934131735,
"eval_loss": 1.0541218519210815,
"eval_runtime": 28.9432,
"eval_samples_per_second": 173.098,
"eval_steps_per_second": 21.663,
"step": 31000
},
{
"epoch": 1.3,
"learning_rate": 8.885277777777779e-07,
"loss": 1.0612,
"step": 32000
},
{
"epoch": 1.3,
"eval_accuracy": 0.4285429141716567,
"eval_loss": 1.0527743101119995,
"eval_runtime": 29.0895,
"eval_samples_per_second": 172.227,
"eval_steps_per_second": 21.554,
"step": 32000
},
{
"epoch": 1.34,
"learning_rate": 9.163055555555556e-07,
"loss": 1.0622,
"step": 33000
},
{
"epoch": 1.34,
"eval_accuracy": 0.4339321357285429,
"eval_loss": 1.0462977886199951,
"eval_runtime": 29.1293,
"eval_samples_per_second": 171.992,
"eval_steps_per_second": 21.525,
"step": 33000
},
{
"epoch": 1.39,
"learning_rate": 9.440555555555557e-07,
"loss": 1.0542,
"step": 34000
},
{
"epoch": 1.39,
"eval_accuracy": 0.43313373253493015,
"eval_loss": 1.044585943222046,
"eval_runtime": 28.9734,
"eval_samples_per_second": 172.917,
"eval_steps_per_second": 21.641,
"step": 34000
},
{
"epoch": 1.43,
"learning_rate": 9.718333333333334e-07,
"loss": 1.0588,
"step": 35000
},
{
"epoch": 1.43,
"eval_accuracy": 0.4281437125748503,
"eval_loss": 1.0411622524261475,
"eval_runtime": 28.9957,
"eval_samples_per_second": 172.784,
"eval_steps_per_second": 21.624,
"step": 35000
},
{
"epoch": 1.47,
"learning_rate": 9.995833333333334e-07,
"loss": 1.0556,
"step": 36000
},
{
"epoch": 1.47,
"eval_accuracy": 0.44471057884231535,
"eval_loss": 1.037837266921997,
"eval_runtime": 29.0336,
"eval_samples_per_second": 172.559,
"eval_steps_per_second": 21.596,
"step": 36000
},
{
"epoch": 1.51,
"learning_rate": 1.0273611111111112e-06,
"loss": 1.0548,
"step": 37000
},
{
"epoch": 1.51,
"eval_accuracy": 0.4502994011976048,
"eval_loss": 1.035124659538269,
"eval_runtime": 29.0914,
"eval_samples_per_second": 172.216,
"eval_steps_per_second": 21.553,
"step": 37000
},
{
"epoch": 1.55,
"learning_rate": 1.055138888888889e-06,
"loss": 1.0566,
"step": 38000
},
{
"epoch": 1.55,
"eval_accuracy": 0.4497005988023952,
"eval_loss": 1.032871127128601,
"eval_runtime": 29.0304,
"eval_samples_per_second": 172.578,
"eval_steps_per_second": 21.598,
"step": 38000
},
{
"epoch": 1.59,
"learning_rate": 1.0829166666666667e-06,
"loss": 1.0513,
"step": 39000
},
{
"epoch": 1.59,
"eval_accuracy": 0.4477045908183633,
"eval_loss": 1.0365269184112549,
"eval_runtime": 29.0199,
"eval_samples_per_second": 172.64,
"eval_steps_per_second": 21.606,
"step": 39000
},
{
"epoch": 1.63,
"learning_rate": 1.1106666666666668e-06,
"loss": 1.0453,
"step": 40000
},
{
"epoch": 1.63,
"eval_accuracy": 0.45708582834331335,
"eval_loss": 1.0288387537002563,
"eval_runtime": 29.1526,
"eval_samples_per_second": 171.854,
"eval_steps_per_second": 21.508,
"step": 40000
},
{
"epoch": 1.67,
"learning_rate": 1.1384166666666669e-06,
"loss": 1.048,
"step": 41000
},
{
"epoch": 1.67,
"eval_accuracy": 0.44530938123752495,
"eval_loss": 1.0338605642318726,
"eval_runtime": 29.1871,
"eval_samples_per_second": 171.651,
"eval_steps_per_second": 21.482,
"step": 41000
},
{
"epoch": 1.71,
"learning_rate": 1.1661944444444447e-06,
"loss": 1.042,
"step": 42000
},
{
"epoch": 1.71,
"eval_accuracy": 0.45109780439121755,
"eval_loss": 1.0249321460723877,
"eval_runtime": 29.0275,
"eval_samples_per_second": 172.595,
"eval_steps_per_second": 21.6,
"step": 42000
},
{
"epoch": 1.75,
"learning_rate": 1.1939444444444445e-06,
"loss": 1.0437,
"step": 43000
},
{
"epoch": 1.75,
"eval_accuracy": 0.44510978043912175,
"eval_loss": 1.0416841506958008,
"eval_runtime": 29.0578,
"eval_samples_per_second": 172.415,
"eval_steps_per_second": 21.578,
"step": 43000
},
{
"epoch": 1.79,
"learning_rate": 1.2217222222222223e-06,
"loss": 1.0454,
"step": 44000
},
{
"epoch": 1.79,
"eval_accuracy": 0.4630738522954092,
"eval_loss": 1.020729422569275,
"eval_runtime": 29.0874,
"eval_samples_per_second": 172.24,
"eval_steps_per_second": 21.556,
"step": 44000
},
{
"epoch": 1.83,
"learning_rate": 1.2495e-06,
"loss": 1.0439,
"step": 45000
},
{
"epoch": 1.83,
"eval_accuracy": 0.4746506986027944,
"eval_loss": 1.0175539255142212,
"eval_runtime": 29.093,
"eval_samples_per_second": 172.206,
"eval_steps_per_second": 21.552,
"step": 45000
},
{
"epoch": 1.87,
"learning_rate": 1.2772500000000001e-06,
"loss": 1.0412,
"step": 46000
},
{
"epoch": 1.87,
"eval_accuracy": 0.4626746506986028,
"eval_loss": 1.0218861103057861,
"eval_runtime": 28.9933,
"eval_samples_per_second": 172.798,
"eval_steps_per_second": 21.626,
"step": 46000
},
{
"epoch": 1.91,
"learning_rate": 1.3050277777777777e-06,
"loss": 1.0362,
"step": 47000
},
{
"epoch": 1.91,
"eval_accuracy": 0.47944111776447107,
"eval_loss": 1.0074430704116821,
"eval_runtime": 29.1004,
"eval_samples_per_second": 172.163,
"eval_steps_per_second": 21.546,
"step": 47000
},
{
"epoch": 1.96,
"learning_rate": 1.3328055555555555e-06,
"loss": 1.0391,
"step": 48000
},
{
"epoch": 1.96,
"eval_accuracy": 0.4614770459081836,
"eval_loss": 1.0313252210617065,
"eval_runtime": 29.0173,
"eval_samples_per_second": 172.655,
"eval_steps_per_second": 21.608,
"step": 48000
},
{
"epoch": 2.0,
"learning_rate": 1.3605555555555555e-06,
"loss": 1.0331,
"step": 49000
},
{
"epoch": 2.0,
"eval_accuracy": 0.48303393213572854,
"eval_loss": 1.0042223930358887,
"eval_runtime": 29.0485,
"eval_samples_per_second": 172.47,
"eval_steps_per_second": 21.585,
"step": 49000
},
{
"epoch": 2.04,
"learning_rate": 1.3883333333333333e-06,
"loss": 1.0328,
"step": 50000
},
{
"epoch": 2.04,
"eval_accuracy": 0.4810379241516966,
"eval_loss": 1.0079505443572998,
"eval_runtime": 29.4084,
"eval_samples_per_second": 170.359,
"eval_steps_per_second": 21.32,
"step": 50000
},
{
"epoch": 2.08,
"learning_rate": 1.4160833333333334e-06,
"loss": 1.0325,
"step": 51000
},
{
"epoch": 2.08,
"eval_accuracy": 0.48403193612774453,
"eval_loss": 1.0023722648620605,
"eval_runtime": 29.4825,
"eval_samples_per_second": 169.932,
"eval_steps_per_second": 21.267,
"step": 51000
},
{
"epoch": 2.12,
"learning_rate": 1.4438611111111112e-06,
"loss": 1.029,
"step": 52000
},
{
"epoch": 2.12,
"eval_accuracy": 0.4786427145708583,
"eval_loss": 1.015783667564392,
"eval_runtime": 29.182,
"eval_samples_per_second": 171.681,
"eval_steps_per_second": 21.486,
"step": 52000
},
{
"epoch": 2.16,
"learning_rate": 1.471638888888889e-06,
"loss": 1.0317,
"step": 53000
},
{
"epoch": 2.16,
"eval_accuracy": 0.49001996007984033,
"eval_loss": 0.9999898672103882,
"eval_runtime": 29.4591,
"eval_samples_per_second": 170.067,
"eval_steps_per_second": 21.284,
"step": 53000
},
{
"epoch": 2.2,
"learning_rate": 1.4993888888888888e-06,
"loss": 1.0319,
"step": 54000
},
{
"epoch": 2.2,
"eval_accuracy": 0.48882235528942114,
"eval_loss": 1.0013622045516968,
"eval_runtime": 29.3854,
"eval_samples_per_second": 170.493,
"eval_steps_per_second": 21.337,
"step": 54000
},
{
"epoch": 2.24,
"learning_rate": 1.5271666666666668e-06,
"loss": 1.0269,
"step": 55000
},
{
"epoch": 2.24,
"eval_accuracy": 0.4878243512974052,
"eval_loss": 0.9981361627578735,
"eval_runtime": 29.0692,
"eval_samples_per_second": 172.347,
"eval_steps_per_second": 21.569,
"step": 55000
},
{
"epoch": 2.28,
"learning_rate": 1.5549166666666666e-06,
"loss": 1.0236,
"step": 56000
},
{
"epoch": 2.28,
"eval_accuracy": 0.474251497005988,
"eval_loss": 1.0223956108093262,
"eval_runtime": 29.0989,
"eval_samples_per_second": 172.172,
"eval_steps_per_second": 21.547,
"step": 56000
},
{
"epoch": 2.32,
"learning_rate": 1.5826944444444446e-06,
"loss": 1.022,
"step": 57000
},
{
"epoch": 2.32,
"eval_accuracy": 0.48962075848303394,
"eval_loss": 0.999485969543457,
"eval_runtime": 29.0951,
"eval_samples_per_second": 172.194,
"eval_steps_per_second": 21.55,
"step": 57000
},
{
"epoch": 2.36,
"learning_rate": 1.6104444444444447e-06,
"loss": 1.0259,
"step": 58000
},
{
"epoch": 2.36,
"eval_accuracy": 0.49161676646706587,
"eval_loss": 1.00543212890625,
"eval_runtime": 29.0186,
"eval_samples_per_second": 172.648,
"eval_steps_per_second": 21.607,
"step": 58000
},
{
"epoch": 2.4,
"learning_rate": 1.6382222222222222e-06,
"loss": 1.0205,
"step": 59000
},
{
"epoch": 2.4,
"eval_accuracy": 0.49620758483033933,
"eval_loss": 0.9923607110977173,
"eval_runtime": 29.1145,
"eval_samples_per_second": 172.079,
"eval_steps_per_second": 21.536,
"step": 59000
},
{
"epoch": 2.44,
"learning_rate": 1.6659722222222223e-06,
"loss": 1.0272,
"step": 60000
},
{
"epoch": 2.44,
"eval_accuracy": 0.49620758483033933,
"eval_loss": 0.9994486570358276,
"eval_runtime": 29.0247,
"eval_samples_per_second": 172.612,
"eval_steps_per_second": 21.602,
"step": 60000
},
{
"epoch": 2.49,
"learning_rate": 1.6937500000000003e-06,
"loss": 1.023,
"step": 61000
},
{
"epoch": 2.49,
"eval_accuracy": 0.49500998003992014,
"eval_loss": 0.9992738366127014,
"eval_runtime": 29.024,
"eval_samples_per_second": 172.616,
"eval_steps_per_second": 21.603,
"step": 61000
},
{
"epoch": 2.53,
"learning_rate": 1.7215000000000002e-06,
"loss": 1.0257,
"step": 62000
},
{
"epoch": 2.53,
"eval_accuracy": 0.481437125748503,
"eval_loss": 1.0081721544265747,
"eval_runtime": 29.125,
"eval_samples_per_second": 172.017,
"eval_steps_per_second": 21.528,
"step": 62000
},
{
"epoch": 2.57,
"learning_rate": 1.7492777777777777e-06,
"loss": 1.0233,
"step": 63000
},
{
"epoch": 2.57,
"eval_accuracy": 0.49660678642714573,
"eval_loss": 0.9943819642066956,
"eval_runtime": 29.3033,
"eval_samples_per_second": 170.97,
"eval_steps_per_second": 21.397,
"step": 63000
},
{
"epoch": 2.61,
"learning_rate": 1.7770277777777778e-06,
"loss": 1.0207,
"step": 64000
},
{
"epoch": 2.61,
"eval_accuracy": 0.500998003992016,
"eval_loss": 0.9942530989646912,
"eval_runtime": 29.3586,
"eval_samples_per_second": 170.648,
"eval_steps_per_second": 21.357,
"step": 64000
},
{
"epoch": 2.65,
"learning_rate": 1.8048055555555558e-06,
"loss": 1.0212,
"step": 65000
},
{
"epoch": 2.65,
"eval_accuracy": 0.5,
"eval_loss": 0.996075451374054,
"eval_runtime": 29.5902,
"eval_samples_per_second": 169.313,
"eval_steps_per_second": 21.189,
"step": 65000
},
{
"epoch": 2.69,
"learning_rate": 1.8325833333333333e-06,
"loss": 1.0259,
"step": 66000
},
{
"epoch": 2.69,
"eval_accuracy": 0.5043912175648703,
"eval_loss": 0.991050124168396,
"eval_runtime": 29.5206,
"eval_samples_per_second": 169.712,
"eval_steps_per_second": 21.239,
"step": 66000
},
{
"epoch": 2.73,
"learning_rate": 1.8603333333333334e-06,
"loss": 1.0199,
"step": 67000
},
{
"epoch": 2.73,
"eval_accuracy": 0.49181636726546907,
"eval_loss": 0.9968613386154175,
"eval_runtime": 29.1614,
"eval_samples_per_second": 171.802,
"eval_steps_per_second": 21.501,
"step": 67000
},
{
"epoch": 2.77,
"learning_rate": 1.8881111111111114e-06,
"loss": 1.0215,
"step": 68000
},
{
"epoch": 2.77,
"eval_accuracy": 0.5023952095808383,
"eval_loss": 0.9903049468994141,
"eval_runtime": 29.1422,
"eval_samples_per_second": 171.916,
"eval_steps_per_second": 21.515,
"step": 68000
},
{
"epoch": 2.81,
"learning_rate": 1.9158611111111115e-06,
"loss": 1.0189,
"step": 69000
},
{
"epoch": 2.81,
"eval_accuracy": 0.5083832335329341,
"eval_loss": 0.9923155903816223,
"eval_runtime": 29.1108,
"eval_samples_per_second": 172.101,
"eval_steps_per_second": 21.538,
"step": 69000
},
{
"epoch": 2.85,
"learning_rate": 1.943638888888889e-06,
"loss": 1.02,
"step": 70000
},
{
"epoch": 2.85,
"eval_accuracy": 0.4992015968063872,
"eval_loss": 0.99754798412323,
"eval_runtime": 29.0673,
"eval_samples_per_second": 172.359,
"eval_steps_per_second": 21.571,
"step": 70000
},
{
"epoch": 2.89,
"learning_rate": 1.971388888888889e-06,
"loss": 1.0124,
"step": 71000
},
{
"epoch": 2.89,
"eval_accuracy": 0.5003992015968064,
"eval_loss": 0.9986121654510498,
"eval_runtime": 29.1291,
"eval_samples_per_second": 171.993,
"eval_steps_per_second": 21.525,
"step": 71000
},
{
"epoch": 2.93,
"learning_rate": 1.999166666666667e-06,
"loss": 1.0212,
"step": 72000
},
{
"epoch": 2.93,
"eval_accuracy": 0.48862275449101794,
"eval_loss": 1.0082526206970215,
"eval_runtime": 29.139,
"eval_samples_per_second": 171.935,
"eval_steps_per_second": 21.518,
"step": 72000
},
{
"epoch": 2.97,
"learning_rate": 2.0269166666666667e-06,
"loss": 1.019,
"step": 73000
},
{
"epoch": 2.97,
"eval_accuracy": 0.5025948103792415,
"eval_loss": 0.9981805682182312,
"eval_runtime": 29.4021,
"eval_samples_per_second": 170.396,
"eval_steps_per_second": 21.325,
"step": 73000
},
{
"epoch": 3.01,
"learning_rate": 2.0546944444444447e-06,
"loss": 1.0205,
"step": 74000
},
{
"epoch": 3.01,
"eval_accuracy": 0.49800399201596807,
"eval_loss": 0.9870504140853882,
"eval_runtime": 29.1143,
"eval_samples_per_second": 172.081,
"eval_steps_per_second": 21.536,
"step": 74000
},
{
"epoch": 3.06,
"learning_rate": 2.0824444444444446e-06,
"loss": 1.0085,
"step": 75000
},
{
"epoch": 3.06,
"eval_accuracy": 0.5017964071856288,
"eval_loss": 0.9926252961158752,
"eval_runtime": 29.2368,
"eval_samples_per_second": 171.359,
"eval_steps_per_second": 21.446,
"step": 75000
},
{
"epoch": 3.1,
"learning_rate": 2.1102222222222226e-06,
"loss": 1.0116,
"step": 76000
},
{
"epoch": 3.1,
"eval_accuracy": 0.5039920159680639,
"eval_loss": 0.9843892455101013,
"eval_runtime": 29.4161,
"eval_samples_per_second": 170.315,
"eval_steps_per_second": 21.315,
"step": 76000
},
{
"epoch": 3.14,
"learning_rate": 2.1379722222222224e-06,
"loss": 1.0128,
"step": 77000
},
{
"epoch": 3.14,
"eval_accuracy": 0.4746506986027944,
"eval_loss": 1.0196081399917603,
"eval_runtime": 29.2722,
"eval_samples_per_second": 171.152,
"eval_steps_per_second": 21.42,
"step": 77000
},
{
"epoch": 3.18,
"learning_rate": 2.16575e-06,
"loss": 1.0051,
"step": 78000
},
{
"epoch": 3.18,
"eval_accuracy": 0.5039920159680639,
"eval_loss": 0.9861226677894592,
"eval_runtime": 29.0739,
"eval_samples_per_second": 172.32,
"eval_steps_per_second": 21.566,
"step": 78000
},
{
"epoch": 3.22,
"learning_rate": 2.1935000000000003e-06,
"loss": 1.0106,
"step": 79000
},
{
"epoch": 3.22,
"eval_accuracy": 0.49241516966067866,
"eval_loss": 1.0106168985366821,
"eval_runtime": 29.2182,
"eval_samples_per_second": 171.469,
"eval_steps_per_second": 21.459,
"step": 79000
},
{
"epoch": 3.26,
"learning_rate": 2.221277777777778e-06,
"loss": 1.0086,
"step": 80000
},
{
"epoch": 3.26,
"eval_accuracy": 0.4942115768463074,
"eval_loss": 1.0031169652938843,
"eval_runtime": 29.1384,
"eval_samples_per_second": 171.938,
"eval_steps_per_second": 21.518,
"step": 80000
},
{
"epoch": 3.3,
"learning_rate": 2.249027777777778e-06,
"loss": 1.0097,
"step": 81000
},
{
"epoch": 3.3,
"eval_accuracy": 0.4932135728542914,
"eval_loss": 0.9960971474647522,
"eval_runtime": 29.1317,
"eval_samples_per_second": 171.977,
"eval_steps_per_second": 21.523,
"step": 81000
},
{
"epoch": 3.34,
"learning_rate": 2.2768055555555557e-06,
"loss": 1.0192,
"step": 82000
},
{
"epoch": 3.34,
"eval_accuracy": 0.49620758483033933,
"eval_loss": 0.9951682686805725,
"eval_runtime": 29.1358,
"eval_samples_per_second": 171.954,
"eval_steps_per_second": 21.52,
"step": 82000
},
{
"epoch": 3.38,
"learning_rate": 2.3045555555555555e-06,
"loss": 1.006,
"step": 83000
},
{
"epoch": 3.38,
"eval_accuracy": 0.49840319361277446,
"eval_loss": 0.9937891960144043,
"eval_runtime": 29.1682,
"eval_samples_per_second": 171.762,
"eval_steps_per_second": 21.496,
"step": 83000
},
{
"epoch": 3.42,
"learning_rate": 2.3323333333333335e-06,
"loss": 1.0118,
"step": 84000
},
{
"epoch": 3.42,
"eval_accuracy": 0.5007984031936128,
"eval_loss": 0.9840214252471924,
"eval_runtime": 29.1175,
"eval_samples_per_second": 172.061,
"eval_steps_per_second": 21.533,
"step": 84000
},
{
"epoch": 3.46,
"learning_rate": 2.360083333333334e-06,
"loss": 1.006,
"step": 85000
},
{
"epoch": 3.46,
"eval_accuracy": 0.49620758483033933,
"eval_loss": 0.9880582690238953,
"eval_runtime": 29.1304,
"eval_samples_per_second": 171.985,
"eval_steps_per_second": 21.524,
"step": 85000
},
{
"epoch": 3.5,
"learning_rate": 2.3878611111111113e-06,
"loss": 1.0151,
"step": 86000
},
{
"epoch": 3.5,
"eval_accuracy": 0.5053892215568863,
"eval_loss": 0.9828479886054993,
"eval_runtime": 29.1321,
"eval_samples_per_second": 171.975,
"eval_steps_per_second": 21.523,
"step": 86000
},
{
"epoch": 3.54,
"learning_rate": 2.4156388888888893e-06,
"loss": 1.0028,
"step": 87000
},
{
"epoch": 3.54,
"eval_accuracy": 0.5073852295409181,
"eval_loss": 0.9821121692657471,
"eval_runtime": 29.1459,
"eval_samples_per_second": 171.894,
"eval_steps_per_second": 21.512,
"step": 87000
},
{
"epoch": 3.59,
"learning_rate": 2.443388888888889e-06,
"loss": 1.0092,
"step": 88000
},
{
"epoch": 3.59,
"eval_accuracy": 0.5105788423153692,
"eval_loss": 0.9791818857192993,
"eval_runtime": 29.1206,
"eval_samples_per_second": 172.043,
"eval_steps_per_second": 21.531,
"step": 88000
},
{
"epoch": 3.63,
"learning_rate": 2.4711666666666668e-06,
"loss": 1.0024,
"step": 89000
},
{
"epoch": 3.63,
"eval_accuracy": 0.4940119760479042,
"eval_loss": 0.986344039440155,
"eval_runtime": 29.1948,
"eval_samples_per_second": 171.606,
"eval_steps_per_second": 21.476,
"step": 89000
},
{
"epoch": 3.67,
"learning_rate": 2.4989166666666666e-06,
"loss": 1.0112,
"step": 90000
},
{
"epoch": 3.67,
"eval_accuracy": 0.5101796407185629,
"eval_loss": 0.9789302349090576,
"eval_runtime": 29.2661,
"eval_samples_per_second": 171.188,
"eval_steps_per_second": 21.424,
"step": 90000
},
{
"epoch": 3.71,
"learning_rate": 2.5266944444444446e-06,
"loss": 1.0137,
"step": 91000
},
{
"epoch": 3.71,
"eval_accuracy": 0.4936127744510978,
"eval_loss": 0.9870284795761108,
"eval_runtime": 29.4012,
"eval_samples_per_second": 170.401,
"eval_steps_per_second": 21.326,
"step": 91000
},
{
"epoch": 3.75,
"learning_rate": 2.554444444444445e-06,
"loss": 1.0076,
"step": 92000
},
{
"epoch": 3.75,
"eval_accuracy": 0.49880239520958086,
"eval_loss": 0.9906250238418579,
"eval_runtime": 29.2222,
"eval_samples_per_second": 171.445,
"eval_steps_per_second": 21.456,
"step": 92000
},
{
"epoch": 3.79,
"learning_rate": 2.5822222222222224e-06,
"loss": 1.0093,
"step": 93000
},
{
"epoch": 3.79,
"eval_accuracy": 0.5121756487025948,
"eval_loss": 0.9766275882720947,
"eval_runtime": 29.2374,
"eval_samples_per_second": 171.356,
"eval_steps_per_second": 21.445,
"step": 93000
},
{
"epoch": 3.83,
"learning_rate": 2.6099722222222223e-06,
"loss": 1.0015,
"step": 94000
},
{
"epoch": 3.83,
"eval_accuracy": 0.4942115768463074,
"eval_loss": 0.991917073726654,
"eval_runtime": 29.3042,
"eval_samples_per_second": 170.965,
"eval_steps_per_second": 21.396,
"step": 94000
},
{
"epoch": 3.87,
"learning_rate": 2.6377500000000003e-06,
"loss": 1.0066,
"step": 95000
},
{
"epoch": 3.87,
"eval_accuracy": 0.5101796407185629,
"eval_loss": 0.9762089252471924,
"eval_runtime": 29.2686,
"eval_samples_per_second": 171.173,
"eval_steps_per_second": 21.422,
"step": 95000
},
{
"epoch": 3.91,
"learning_rate": 2.6655e-06,
"loss": 1.0033,
"step": 96000
},
{
"epoch": 3.91,
"eval_accuracy": 0.5071856287425149,
"eval_loss": 0.9739634990692139,
"eval_runtime": 29.2989,
"eval_samples_per_second": 170.996,
"eval_steps_per_second": 21.4,
"step": 96000
},
{
"epoch": 3.95,
"learning_rate": 2.6932777777777777e-06,
"loss": 1.0054,
"step": 97000
},
{
"epoch": 3.95,
"eval_accuracy": 0.5033932135728543,
"eval_loss": 0.9856627583503723,
"eval_runtime": 29.2684,
"eval_samples_per_second": 171.174,
"eval_steps_per_second": 21.422,
"step": 97000
},
{
"epoch": 3.99,
"learning_rate": 2.721027777777778e-06,
"loss": 1.0034,
"step": 98000
},
{
"epoch": 3.99,
"eval_accuracy": 0.5055888223552895,
"eval_loss": 0.9802790880203247,
"eval_runtime": 29.2896,
"eval_samples_per_second": 171.051,
"eval_steps_per_second": 21.407,
"step": 98000
},
{
"epoch": 4.03,
"learning_rate": 2.748805555555556e-06,
"loss": 1.0003,
"step": 99000
},
{
"epoch": 4.03,
"eval_accuracy": 0.5183632734530939,
"eval_loss": 0.9739055037498474,
"eval_runtime": 29.3152,
"eval_samples_per_second": 170.901,
"eval_steps_per_second": 21.388,
"step": 99000
},
{
"epoch": 4.07,
"learning_rate": 2.776555555555556e-06,
"loss": 0.9978,
"step": 100000
},
{
"epoch": 4.07,
"eval_accuracy": 0.5153692614770459,
"eval_loss": 0.9735556244850159,
"eval_runtime": 29.6841,
"eval_samples_per_second": 168.777,
"eval_steps_per_second": 21.122,
"step": 100000
},
{
"epoch": 4.12,
"learning_rate": 2.8043333333333334e-06,
"loss": 1.0012,
"step": 101000
},
{
"epoch": 4.12,
"eval_accuracy": 0.5115768463073852,
"eval_loss": 0.9762488603591919,
"eval_runtime": 29.3408,
"eval_samples_per_second": 170.752,
"eval_steps_per_second": 21.37,
"step": 101000
},
{
"epoch": 4.16,
"learning_rate": 2.8320833333333333e-06,
"loss": 0.9995,
"step": 102000
},
{
"epoch": 4.16,
"eval_accuracy": 0.492814371257485,
"eval_loss": 0.986444890499115,
"eval_runtime": 29.2787,
"eval_samples_per_second": 171.114,
"eval_steps_per_second": 21.415,
"step": 102000
},
{
"epoch": 4.2,
"learning_rate": 2.8598611111111112e-06,
"loss": 1.0034,
"step": 103000
},
{
"epoch": 4.2,
"eval_accuracy": 0.5181636726546907,
"eval_loss": 0.9686606526374817,
"eval_runtime": 29.3082,
"eval_samples_per_second": 170.942,
"eval_steps_per_second": 21.393,
"step": 103000
},
{
"epoch": 4.24,
"learning_rate": 2.8876111111111115e-06,
"loss": 0.9966,
"step": 104000
},
{
"epoch": 4.24,
"eval_accuracy": 0.49540918163672654,
"eval_loss": 0.9904249906539917,
"eval_runtime": 29.3774,
"eval_samples_per_second": 170.539,
"eval_steps_per_second": 21.343,
"step": 104000
},
{
"epoch": 4.28,
"learning_rate": 2.915388888888889e-06,
"loss": 0.993,
"step": 105000
},
{
"epoch": 4.28,
"eval_accuracy": 0.48522954091816367,
"eval_loss": 1.0059425830841064,
"eval_runtime": 29.3509,
"eval_samples_per_second": 170.693,
"eval_steps_per_second": 21.362,
"step": 105000
},
{
"epoch": 4.32,
"learning_rate": 2.943138888888889e-06,
"loss": 0.9887,
"step": 106000
},
{
"epoch": 4.32,
"eval_accuracy": 0.5167664670658683,
"eval_loss": 0.9730390906333923,
"eval_runtime": 29.3622,
"eval_samples_per_second": 170.628,
"eval_steps_per_second": 21.354,
"step": 106000
},
{
"epoch": 4.36,
"learning_rate": 2.9709166666666665e-06,
"loss": 0.99,
"step": 107000
},
{
"epoch": 4.36,
"eval_accuracy": 0.5067864271457085,
"eval_loss": 0.9823795557022095,
"eval_runtime": 29.6253,
"eval_samples_per_second": 169.112,
"eval_steps_per_second": 21.164,
"step": 107000
},
{
"epoch": 4.4,
"learning_rate": 2.9986666666666668e-06,
"loss": 0.9892,
"step": 108000
},
{
"epoch": 4.4,
"eval_accuracy": 0.5063872255489021,
"eval_loss": 0.9779866337776184,
"eval_runtime": 29.3617,
"eval_samples_per_second": 170.63,
"eval_steps_per_second": 21.354,
"step": 108000
},
{
"epoch": 4.44,
"learning_rate": 3.0264444444444448e-06,
"loss": 0.9875,
"step": 109000
},
{
"epoch": 4.44,
"eval_accuracy": 0.51437125748503,
"eval_loss": 0.9730616211891174,
"eval_runtime": 29.1847,
"eval_samples_per_second": 171.665,
"eval_steps_per_second": 21.484,
"step": 109000
},
{
"epoch": 4.48,
"learning_rate": 3.0541944444444446e-06,
"loss": 0.9941,
"step": 110000
},
{
"epoch": 4.48,
"eval_accuracy": 0.5141716566866268,
"eval_loss": 0.9669836163520813,
"eval_runtime": 29.1138,
"eval_samples_per_second": 172.084,
"eval_steps_per_second": 21.536,
"step": 110000
},
{
"epoch": 4.52,
"learning_rate": 3.081972222222222e-06,
"loss": 0.9882,
"step": 111000
},
{
"epoch": 4.52,
"eval_accuracy": 0.5111776447105788,
"eval_loss": 0.9662576913833618,
"eval_runtime": 29.1798,
"eval_samples_per_second": 171.694,
"eval_steps_per_second": 21.487,
"step": 111000
},
{
"epoch": 4.56,
"learning_rate": 3.109722222222222e-06,
"loss": 1.0006,
"step": 112000
},
{
"epoch": 4.56,
"eval_accuracy": 0.5165668662674651,
"eval_loss": 0.9604237079620361,
"eval_runtime": 29.1464,
"eval_samples_per_second": 171.891,
"eval_steps_per_second": 21.512,
"step": 112000
},
{
"epoch": 4.6,
"learning_rate": 3.1375e-06,
"loss": 0.9836,
"step": 113000
},
{
"epoch": 4.6,
"eval_accuracy": 0.5065868263473053,
"eval_loss": 0.9700830578804016,
"eval_runtime": 29.1062,
"eval_samples_per_second": 172.128,
"eval_steps_per_second": 21.542,
"step": 113000
}
],
"logging_steps": 1000,
"max_steps": 10000000,
"num_train_epochs": 408,
"save_steps": 1000,
"total_flos": 1.1810774146822963e+17,
"trial_name": null,
"trial_params": null
}