tmnam20's picture
Training in progress, step 111000, checkpoint
a808dad
{
"best_metric": 0.9607908129692078,
"best_model_checkpoint": "/kaggle/output/checkpoint-111000",
"epoch": 4.522490221642764,
"eval_steps": 1000,
"global_step": 111000,
"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": 1.05,
"step": 1
},
{
"epoch": 0.04,
"learning_rate": 2.7750000000000004e-08,
"loss": 1.1403,
"step": 1000
},
{
"epoch": 0.04,
"eval_accuracy": 0.33712574850299404,
"eval_loss": 1.1030837297439575,
"eval_runtime": 29.7882,
"eval_samples_per_second": 168.187,
"eval_steps_per_second": 21.049,
"step": 1000
},
{
"epoch": 0.08,
"learning_rate": 5.5527777777777784e-08,
"loss": 1.1194,
"step": 2000
},
{
"epoch": 0.08,
"eval_accuracy": 0.34331337325349304,
"eval_loss": 1.1002165079116821,
"eval_runtime": 29.8685,
"eval_samples_per_second": 167.735,
"eval_steps_per_second": 20.992,
"step": 2000
},
{
"epoch": 0.12,
"learning_rate": 8.327777777777778e-08,
"loss": 1.1194,
"step": 3000
},
{
"epoch": 0.12,
"eval_accuracy": 0.3471057884231537,
"eval_loss": 1.101266622543335,
"eval_runtime": 29.709,
"eval_samples_per_second": 168.636,
"eval_steps_per_second": 21.105,
"step": 3000
},
{
"epoch": 0.16,
"learning_rate": 1.1105555555555557e-07,
"loss": 1.1166,
"step": 4000
},
{
"epoch": 0.16,
"eval_accuracy": 0.34311377245508984,
"eval_loss": 1.0984796285629272,
"eval_runtime": 29.9401,
"eval_samples_per_second": 167.334,
"eval_steps_per_second": 20.942,
"step": 4000
},
{
"epoch": 0.2,
"learning_rate": 1.3883333333333335e-07,
"loss": 1.1137,
"step": 5000
},
{
"epoch": 0.2,
"eval_accuracy": 0.3403193612774451,
"eval_loss": 1.0974279642105103,
"eval_runtime": 29.8185,
"eval_samples_per_second": 168.016,
"eval_steps_per_second": 21.027,
"step": 5000
},
{
"epoch": 0.24,
"learning_rate": 1.6658333333333335e-07,
"loss": 1.1109,
"step": 6000
},
{
"epoch": 0.24,
"eval_accuracy": 0.3592814371257485,
"eval_loss": 1.0984892845153809,
"eval_runtime": 29.893,
"eval_samples_per_second": 167.598,
"eval_steps_per_second": 20.975,
"step": 6000
},
{
"epoch": 0.29,
"learning_rate": 1.9436111111111112e-07,
"loss": 1.1122,
"step": 7000
},
{
"epoch": 0.29,
"eval_accuracy": 0.36327345309381237,
"eval_loss": 1.0960686206817627,
"eval_runtime": 29.8315,
"eval_samples_per_second": 167.943,
"eval_steps_per_second": 21.018,
"step": 7000
},
{
"epoch": 0.33,
"learning_rate": 2.2213888888888891e-07,
"loss": 1.1091,
"step": 8000
},
{
"epoch": 0.33,
"eval_accuracy": 0.35728542914171657,
"eval_loss": 1.0958523750305176,
"eval_runtime": 29.9494,
"eval_samples_per_second": 167.282,
"eval_steps_per_second": 20.935,
"step": 8000
},
{
"epoch": 0.37,
"learning_rate": 2.4988888888888893e-07,
"loss": 1.1077,
"step": 9000
},
{
"epoch": 0.37,
"eval_accuracy": 0.35528942115768464,
"eval_loss": 1.1003851890563965,
"eval_runtime": 29.9762,
"eval_samples_per_second": 167.133,
"eval_steps_per_second": 20.917,
"step": 9000
},
{
"epoch": 0.41,
"learning_rate": 2.776666666666667e-07,
"loss": 1.1091,
"step": 10000
},
{
"epoch": 0.41,
"eval_accuracy": 0.3790419161676647,
"eval_loss": 1.0951762199401855,
"eval_runtime": 29.8593,
"eval_samples_per_second": 167.787,
"eval_steps_per_second": 20.998,
"step": 10000
},
{
"epoch": 0.45,
"learning_rate": 3.0541666666666667e-07,
"loss": 1.1047,
"step": 11000
},
{
"epoch": 0.45,
"eval_accuracy": 0.3780439121756487,
"eval_loss": 1.0928910970687866,
"eval_runtime": 29.8768,
"eval_samples_per_second": 167.689,
"eval_steps_per_second": 20.986,
"step": 11000
},
{
"epoch": 0.49,
"learning_rate": 3.3319444444444444e-07,
"loss": 1.1053,
"step": 12000
},
{
"epoch": 0.49,
"eval_accuracy": 0.38323353293413176,
"eval_loss": 1.0910215377807617,
"eval_runtime": 29.8703,
"eval_samples_per_second": 167.725,
"eval_steps_per_second": 20.991,
"step": 12000
},
{
"epoch": 0.53,
"learning_rate": 3.6094444444444446e-07,
"loss": 1.1035,
"step": 13000
},
{
"epoch": 0.53,
"eval_accuracy": 0.34850299401197604,
"eval_loss": 1.0930161476135254,
"eval_runtime": 29.8865,
"eval_samples_per_second": 167.634,
"eval_steps_per_second": 20.979,
"step": 13000
},
{
"epoch": 0.57,
"learning_rate": 3.8872222222222223e-07,
"loss": 1.1002,
"step": 14000
},
{
"epoch": 0.57,
"eval_accuracy": 0.3646706586826347,
"eval_loss": 1.093497633934021,
"eval_runtime": 30.0222,
"eval_samples_per_second": 166.877,
"eval_steps_per_second": 20.885,
"step": 14000
},
{
"epoch": 0.61,
"learning_rate": 4.1647222222222225e-07,
"loss": 1.1025,
"step": 15000
},
{
"epoch": 0.61,
"eval_accuracy": 0.37924151696606784,
"eval_loss": 1.0900779962539673,
"eval_runtime": 30.0049,
"eval_samples_per_second": 166.973,
"eval_steps_per_second": 20.897,
"step": 15000
},
{
"epoch": 0.65,
"learning_rate": 4.4425e-07,
"loss": 1.0992,
"step": 16000
},
{
"epoch": 0.65,
"eval_accuracy": 0.40119760479041916,
"eval_loss": 1.0855979919433594,
"eval_runtime": 30.1095,
"eval_samples_per_second": 166.393,
"eval_steps_per_second": 20.824,
"step": 16000
},
{
"epoch": 0.69,
"learning_rate": 4.7200000000000004e-07,
"loss": 1.0965,
"step": 17000
},
{
"epoch": 0.69,
"eval_accuracy": 0.4171656686626746,
"eval_loss": 1.0845381021499634,
"eval_runtime": 30.0586,
"eval_samples_per_second": 166.674,
"eval_steps_per_second": 20.859,
"step": 17000
},
{
"epoch": 0.73,
"learning_rate": 4.997777777777779e-07,
"loss": 1.0962,
"step": 18000
},
{
"epoch": 0.73,
"eval_accuracy": 0.4087824351297405,
"eval_loss": 1.0860552787780762,
"eval_runtime": 30.2343,
"eval_samples_per_second": 165.706,
"eval_steps_per_second": 20.738,
"step": 18000
},
{
"epoch": 0.77,
"learning_rate": 5.275277777777778e-07,
"loss": 1.0936,
"step": 19000
},
{
"epoch": 0.77,
"eval_accuracy": 0.4039920159680639,
"eval_loss": 1.0822259187698364,
"eval_runtime": 30.216,
"eval_samples_per_second": 165.806,
"eval_steps_per_second": 20.751,
"step": 19000
},
{
"epoch": 0.81,
"learning_rate": 5.553055555555556e-07,
"loss": 1.0923,
"step": 20000
},
{
"epoch": 0.81,
"eval_accuracy": 0.4095808383233533,
"eval_loss": 1.0820951461791992,
"eval_runtime": 29.968,
"eval_samples_per_second": 167.178,
"eval_steps_per_second": 20.922,
"step": 20000
},
{
"epoch": 0.86,
"learning_rate": 5.830555555555556e-07,
"loss": 1.0897,
"step": 21000
},
{
"epoch": 0.86,
"eval_accuracy": 0.4149700598802395,
"eval_loss": 1.0763800144195557,
"eval_runtime": 30.021,
"eval_samples_per_second": 166.883,
"eval_steps_per_second": 20.885,
"step": 21000
},
{
"epoch": 0.9,
"learning_rate": 6.108333333333333e-07,
"loss": 1.0888,
"step": 22000
},
{
"epoch": 0.9,
"eval_accuracy": 0.43193612774451096,
"eval_loss": 1.0687414407730103,
"eval_runtime": 30.1577,
"eval_samples_per_second": 166.127,
"eval_steps_per_second": 20.791,
"step": 22000
},
{
"epoch": 0.94,
"learning_rate": 6.385833333333334e-07,
"loss": 1.0823,
"step": 23000
},
{
"epoch": 0.94,
"eval_accuracy": 0.4285429141716567,
"eval_loss": 1.0642070770263672,
"eval_runtime": 30.0701,
"eval_samples_per_second": 166.611,
"eval_steps_per_second": 20.851,
"step": 23000
},
{
"epoch": 0.98,
"learning_rate": 6.663611111111112e-07,
"loss": 1.0786,
"step": 24000
},
{
"epoch": 0.98,
"eval_accuracy": 0.4363273453093812,
"eval_loss": 1.0598841905593872,
"eval_runtime": 30.0689,
"eval_samples_per_second": 166.617,
"eval_steps_per_second": 20.852,
"step": 24000
},
{
"epoch": 1.02,
"learning_rate": 6.941111111111112e-07,
"loss": 1.0708,
"step": 25000
},
{
"epoch": 1.02,
"eval_accuracy": 0.4429141716566866,
"eval_loss": 1.0561795234680176,
"eval_runtime": 30.1301,
"eval_samples_per_second": 166.279,
"eval_steps_per_second": 20.81,
"step": 25000
},
{
"epoch": 1.06,
"learning_rate": 7.218888888888889e-07,
"loss": 1.072,
"step": 26000
},
{
"epoch": 1.06,
"eval_accuracy": 0.4303393213572854,
"eval_loss": 1.0574887990951538,
"eval_runtime": 30.0473,
"eval_samples_per_second": 166.737,
"eval_steps_per_second": 20.867,
"step": 26000
},
{
"epoch": 1.1,
"learning_rate": 7.496666666666667e-07,
"loss": 1.0701,
"step": 27000
},
{
"epoch": 1.1,
"eval_accuracy": 0.4421157684630739,
"eval_loss": 1.055459976196289,
"eval_runtime": 30.1049,
"eval_samples_per_second": 166.418,
"eval_steps_per_second": 20.827,
"step": 27000
},
{
"epoch": 1.14,
"learning_rate": 7.774166666666668e-07,
"loss": 1.0677,
"step": 28000
},
{
"epoch": 1.14,
"eval_accuracy": 0.4285429141716567,
"eval_loss": 1.0527067184448242,
"eval_runtime": 30.1476,
"eval_samples_per_second": 166.182,
"eval_steps_per_second": 20.798,
"step": 28000
},
{
"epoch": 1.18,
"learning_rate": 8.051944444444445e-07,
"loss": 1.065,
"step": 29000
},
{
"epoch": 1.18,
"eval_accuracy": 0.43852295409181635,
"eval_loss": 1.0494154691696167,
"eval_runtime": 30.036,
"eval_samples_per_second": 166.8,
"eval_steps_per_second": 20.875,
"step": 29000
},
{
"epoch": 1.22,
"learning_rate": 8.329444444444445e-07,
"loss": 1.0628,
"step": 30000
},
{
"epoch": 1.22,
"eval_accuracy": 0.4377245508982036,
"eval_loss": 1.0461777448654175,
"eval_runtime": 30.1005,
"eval_samples_per_second": 166.442,
"eval_steps_per_second": 20.83,
"step": 30000
},
{
"epoch": 1.26,
"learning_rate": 8.607222222222223e-07,
"loss": 1.0672,
"step": 31000
},
{
"epoch": 1.26,
"eval_accuracy": 0.4435129740518962,
"eval_loss": 1.0531694889068604,
"eval_runtime": 30.1733,
"eval_samples_per_second": 166.041,
"eval_steps_per_second": 20.78,
"step": 31000
},
{
"epoch": 1.3,
"learning_rate": 8.884722222222224e-07,
"loss": 1.0638,
"step": 32000
},
{
"epoch": 1.3,
"eval_accuracy": 0.4297405189620758,
"eval_loss": 1.0501775741577148,
"eval_runtime": 30.058,
"eval_samples_per_second": 166.678,
"eval_steps_per_second": 20.86,
"step": 32000
},
{
"epoch": 1.34,
"learning_rate": 9.162500000000001e-07,
"loss": 1.0637,
"step": 33000
},
{
"epoch": 1.34,
"eval_accuracy": 0.44491017964071855,
"eval_loss": 1.0452704429626465,
"eval_runtime": 30.3623,
"eval_samples_per_second": 165.007,
"eval_steps_per_second": 20.651,
"step": 33000
},
{
"epoch": 1.39,
"learning_rate": 9.440000000000001e-07,
"loss": 1.0558,
"step": 34000
},
{
"epoch": 1.39,
"eval_accuracy": 0.44510978043912175,
"eval_loss": 1.0434744358062744,
"eval_runtime": 30.2239,
"eval_samples_per_second": 165.763,
"eval_steps_per_second": 20.745,
"step": 34000
},
{
"epoch": 1.43,
"learning_rate": 9.71777777777778e-07,
"loss": 1.0598,
"step": 35000
},
{
"epoch": 1.43,
"eval_accuracy": 0.4473053892215569,
"eval_loss": 1.039801001548767,
"eval_runtime": 30.107,
"eval_samples_per_second": 166.407,
"eval_steps_per_second": 20.826,
"step": 35000
},
{
"epoch": 1.47,
"learning_rate": 9.995277777777778e-07,
"loss": 1.0601,
"step": 36000
},
{
"epoch": 1.47,
"eval_accuracy": 0.4469061876247505,
"eval_loss": 1.037074327468872,
"eval_runtime": 30.1954,
"eval_samples_per_second": 165.919,
"eval_steps_per_second": 20.765,
"step": 36000
},
{
"epoch": 1.51,
"learning_rate": 1.0273055555555556e-06,
"loss": 1.0529,
"step": 37000
},
{
"epoch": 1.51,
"eval_accuracy": 0.45229540918163674,
"eval_loss": 1.0346434116363525,
"eval_runtime": 30.2856,
"eval_samples_per_second": 165.425,
"eval_steps_per_second": 20.703,
"step": 37000
},
{
"epoch": 1.55,
"learning_rate": 1.0550555555555557e-06,
"loss": 1.0544,
"step": 38000
},
{
"epoch": 1.55,
"eval_accuracy": 0.45129740518962075,
"eval_loss": 1.0338674783706665,
"eval_runtime": 30.1511,
"eval_samples_per_second": 166.163,
"eval_steps_per_second": 20.795,
"step": 38000
},
{
"epoch": 1.59,
"learning_rate": 1.0828333333333334e-06,
"loss": 1.0528,
"step": 39000
},
{
"epoch": 1.59,
"eval_accuracy": 0.44550898203592815,
"eval_loss": 1.0376836061477661,
"eval_runtime": 30.304,
"eval_samples_per_second": 165.324,
"eval_steps_per_second": 20.69,
"step": 39000
},
{
"epoch": 1.63,
"learning_rate": 1.1106111111111112e-06,
"loss": 1.0482,
"step": 40000
},
{
"epoch": 1.63,
"eval_accuracy": 0.4540918163672655,
"eval_loss": 1.031366229057312,
"eval_runtime": 30.2062,
"eval_samples_per_second": 165.86,
"eval_steps_per_second": 20.757,
"step": 40000
},
{
"epoch": 1.67,
"learning_rate": 1.1383611111111113e-06,
"loss": 1.0494,
"step": 41000
},
{
"epoch": 1.67,
"eval_accuracy": 0.4552894211576846,
"eval_loss": 1.032992959022522,
"eval_runtime": 30.1667,
"eval_samples_per_second": 166.077,
"eval_steps_per_second": 20.784,
"step": 41000
},
{
"epoch": 1.71,
"learning_rate": 1.166138888888889e-06,
"loss": 1.0474,
"step": 42000
},
{
"epoch": 1.71,
"eval_accuracy": 0.45129740518962075,
"eval_loss": 1.0255250930786133,
"eval_runtime": 30.2122,
"eval_samples_per_second": 165.827,
"eval_steps_per_second": 20.753,
"step": 42000
},
{
"epoch": 1.75,
"learning_rate": 1.193888888888889e-06,
"loss": 1.0448,
"step": 43000
},
{
"epoch": 1.75,
"eval_accuracy": 0.4483033932135729,
"eval_loss": 1.0412585735321045,
"eval_runtime": 30.2083,
"eval_samples_per_second": 165.849,
"eval_steps_per_second": 20.756,
"step": 43000
},
{
"epoch": 1.79,
"learning_rate": 1.2216666666666667e-06,
"loss": 1.0458,
"step": 44000
},
{
"epoch": 1.79,
"eval_accuracy": 0.46427145708582834,
"eval_loss": 1.0211690664291382,
"eval_runtime": 30.4051,
"eval_samples_per_second": 164.775,
"eval_steps_per_second": 20.622,
"step": 44000
},
{
"epoch": 1.83,
"learning_rate": 1.2494166666666668e-06,
"loss": 1.0468,
"step": 45000
},
{
"epoch": 1.83,
"eval_accuracy": 0.47105788423153694,
"eval_loss": 1.0185887813568115,
"eval_runtime": 30.1465,
"eval_samples_per_second": 166.188,
"eval_steps_per_second": 20.798,
"step": 45000
},
{
"epoch": 1.87,
"learning_rate": 1.2771944444444445e-06,
"loss": 1.0414,
"step": 46000
},
{
"epoch": 1.87,
"eval_accuracy": 0.46966067864271455,
"eval_loss": 1.0215678215026855,
"eval_runtime": 30.1615,
"eval_samples_per_second": 166.106,
"eval_steps_per_second": 20.788,
"step": 46000
},
{
"epoch": 1.91,
"learning_rate": 1.3049444444444446e-06,
"loss": 1.0374,
"step": 47000
},
{
"epoch": 1.91,
"eval_accuracy": 0.47684630738522954,
"eval_loss": 1.0100239515304565,
"eval_runtime": 30.0919,
"eval_samples_per_second": 166.49,
"eval_steps_per_second": 20.836,
"step": 47000
},
{
"epoch": 1.96,
"learning_rate": 1.3327222222222224e-06,
"loss": 1.0423,
"step": 48000
},
{
"epoch": 1.96,
"eval_accuracy": 0.462874251497006,
"eval_loss": 1.0277544260025024,
"eval_runtime": 30.0926,
"eval_samples_per_second": 166.486,
"eval_steps_per_second": 20.836,
"step": 48000
},
{
"epoch": 2.0,
"learning_rate": 1.3604722222222224e-06,
"loss": 1.0332,
"step": 49000
},
{
"epoch": 2.0,
"eval_accuracy": 0.4870259481037924,
"eval_loss": 1.0046255588531494,
"eval_runtime": 30.0835,
"eval_samples_per_second": 166.536,
"eval_steps_per_second": 20.842,
"step": 49000
},
{
"epoch": 2.04,
"learning_rate": 1.3882500000000002e-06,
"loss": 1.0337,
"step": 50000
},
{
"epoch": 2.04,
"eval_accuracy": 0.48542914171656687,
"eval_loss": 1.00551438331604,
"eval_runtime": 30.1523,
"eval_samples_per_second": 166.157,
"eval_steps_per_second": 20.794,
"step": 50000
},
{
"epoch": 2.08,
"learning_rate": 1.416e-06,
"loss": 1.0335,
"step": 51000
},
{
"epoch": 2.08,
"eval_accuracy": 0.4782435129740519,
"eval_loss": 1.0022392272949219,
"eval_runtime": 30.0244,
"eval_samples_per_second": 166.864,
"eval_steps_per_second": 20.883,
"step": 51000
},
{
"epoch": 2.12,
"learning_rate": 1.4437777777777779e-06,
"loss": 1.0272,
"step": 52000
},
{
"epoch": 2.12,
"eval_accuracy": 0.48003992015968067,
"eval_loss": 1.014728307723999,
"eval_runtime": 30.0497,
"eval_samples_per_second": 166.724,
"eval_steps_per_second": 20.865,
"step": 52000
},
{
"epoch": 2.16,
"learning_rate": 1.471527777777778e-06,
"loss": 1.0315,
"step": 53000
},
{
"epoch": 2.16,
"eval_accuracy": 0.4848303393213573,
"eval_loss": 0.9981946349143982,
"eval_runtime": 30.2291,
"eval_samples_per_second": 165.734,
"eval_steps_per_second": 20.742,
"step": 53000
},
{
"epoch": 2.2,
"learning_rate": 1.4993055555555557e-06,
"loss": 1.0291,
"step": 54000
},
{
"epoch": 2.2,
"eval_accuracy": 0.49021956087824353,
"eval_loss": 0.9988633990287781,
"eval_runtime": 30.0961,
"eval_samples_per_second": 166.467,
"eval_steps_per_second": 20.833,
"step": 54000
},
{
"epoch": 2.24,
"learning_rate": 1.5270555555555558e-06,
"loss": 1.027,
"step": 55000
},
{
"epoch": 2.24,
"eval_accuracy": 0.48602794411177647,
"eval_loss": 0.99713534116745,
"eval_runtime": 30.1599,
"eval_samples_per_second": 166.115,
"eval_steps_per_second": 20.789,
"step": 55000
},
{
"epoch": 2.28,
"learning_rate": 1.5548333333333335e-06,
"loss": 1.0215,
"step": 56000
},
{
"epoch": 2.28,
"eval_accuracy": 0.46487025948103794,
"eval_loss": 1.0212756395339966,
"eval_runtime": 30.2641,
"eval_samples_per_second": 165.543,
"eval_steps_per_second": 20.718,
"step": 56000
},
{
"epoch": 2.32,
"learning_rate": 1.5825833333333334e-06,
"loss": 1.0203,
"step": 57000
},
{
"epoch": 2.32,
"eval_accuracy": 0.4874251497005988,
"eval_loss": 0.9981949925422668,
"eval_runtime": 30.1111,
"eval_samples_per_second": 166.384,
"eval_steps_per_second": 20.823,
"step": 57000
},
{
"epoch": 2.36,
"learning_rate": 1.6103611111111112e-06,
"loss": 1.0238,
"step": 58000
},
{
"epoch": 2.36,
"eval_accuracy": 0.48562874251497007,
"eval_loss": 1.0043253898620605,
"eval_runtime": 30.4057,
"eval_samples_per_second": 164.771,
"eval_steps_per_second": 20.621,
"step": 58000
},
{
"epoch": 2.4,
"learning_rate": 1.638111111111111e-06,
"loss": 1.0211,
"step": 59000
},
{
"epoch": 2.4,
"eval_accuracy": 0.49820359281437127,
"eval_loss": 0.9920447468757629,
"eval_runtime": 30.4581,
"eval_samples_per_second": 164.488,
"eval_steps_per_second": 20.586,
"step": 59000
},
{
"epoch": 2.44,
"learning_rate": 1.665888888888889e-06,
"loss": 1.0236,
"step": 60000
},
{
"epoch": 2.44,
"eval_accuracy": 0.49241516966067866,
"eval_loss": 0.9953387975692749,
"eval_runtime": 30.3298,
"eval_samples_per_second": 165.184,
"eval_steps_per_second": 20.673,
"step": 60000
},
{
"epoch": 2.49,
"learning_rate": 1.693638888888889e-06,
"loss": 1.0207,
"step": 61000
},
{
"epoch": 2.49,
"eval_accuracy": 0.49760479041916167,
"eval_loss": 0.9943307042121887,
"eval_runtime": 30.3886,
"eval_samples_per_second": 164.865,
"eval_steps_per_second": 20.633,
"step": 61000
},
{
"epoch": 2.53,
"learning_rate": 1.7214166666666666e-06,
"loss": 1.0222,
"step": 62000
},
{
"epoch": 2.53,
"eval_accuracy": 0.4880239520958084,
"eval_loss": 1.005075454711914,
"eval_runtime": 30.41,
"eval_samples_per_second": 164.748,
"eval_steps_per_second": 20.618,
"step": 62000
},
{
"epoch": 2.57,
"learning_rate": 1.7491666666666667e-06,
"loss": 1.0221,
"step": 63000
},
{
"epoch": 2.57,
"eval_accuracy": 0.5011976047904192,
"eval_loss": 0.9902356266975403,
"eval_runtime": 30.3384,
"eval_samples_per_second": 165.137,
"eval_steps_per_second": 20.667,
"step": 63000
},
{
"epoch": 2.61,
"learning_rate": 1.7769444444444447e-06,
"loss": 1.0188,
"step": 64000
},
{
"epoch": 2.61,
"eval_accuracy": 0.5003992015968064,
"eval_loss": 0.9907160997390747,
"eval_runtime": 30.3592,
"eval_samples_per_second": 165.024,
"eval_steps_per_second": 20.653,
"step": 64000
},
{
"epoch": 2.65,
"learning_rate": 1.8046944444444446e-06,
"loss": 1.0187,
"step": 65000
},
{
"epoch": 2.65,
"eval_accuracy": 0.49820359281437127,
"eval_loss": 0.9914441704750061,
"eval_runtime": 30.3614,
"eval_samples_per_second": 165.012,
"eval_steps_per_second": 20.651,
"step": 65000
},
{
"epoch": 2.69,
"learning_rate": 1.8324722222222223e-06,
"loss": 1.023,
"step": 66000
},
{
"epoch": 2.69,
"eval_accuracy": 0.5007984031936128,
"eval_loss": 0.9877445101737976,
"eval_runtime": 30.341,
"eval_samples_per_second": 165.123,
"eval_steps_per_second": 20.665,
"step": 66000
},
{
"epoch": 2.73,
"learning_rate": 1.8602222222222222e-06,
"loss": 1.0195,
"step": 67000
},
{
"epoch": 2.73,
"eval_accuracy": 0.49660678642714573,
"eval_loss": 0.9916561245918274,
"eval_runtime": 30.3344,
"eval_samples_per_second": 165.159,
"eval_steps_per_second": 20.67,
"step": 67000
},
{
"epoch": 2.77,
"learning_rate": 1.8880000000000002e-06,
"loss": 1.0181,
"step": 68000
},
{
"epoch": 2.77,
"eval_accuracy": 0.4992015968063872,
"eval_loss": 0.985427975654602,
"eval_runtime": 30.3261,
"eval_samples_per_second": 165.204,
"eval_steps_per_second": 20.675,
"step": 68000
},
{
"epoch": 2.81,
"learning_rate": 1.9157500000000003e-06,
"loss": 1.0183,
"step": 69000
},
{
"epoch": 2.81,
"eval_accuracy": 0.5033932135728543,
"eval_loss": 0.9862416982650757,
"eval_runtime": 30.3869,
"eval_samples_per_second": 164.874,
"eval_steps_per_second": 20.634,
"step": 69000
},
{
"epoch": 2.85,
"learning_rate": 1.943527777777778e-06,
"loss": 1.0176,
"step": 70000
},
{
"epoch": 2.85,
"eval_accuracy": 0.501996007984032,
"eval_loss": 0.9887968897819519,
"eval_runtime": 30.3292,
"eval_samples_per_second": 165.187,
"eval_steps_per_second": 20.673,
"step": 70000
},
{
"epoch": 2.89,
"learning_rate": 1.9712777777777777e-06,
"loss": 1.0108,
"step": 71000
},
{
"epoch": 2.89,
"eval_accuracy": 0.49520958083832334,
"eval_loss": 0.992777407169342,
"eval_runtime": 30.4,
"eval_samples_per_second": 164.803,
"eval_steps_per_second": 20.625,
"step": 71000
},
{
"epoch": 2.93,
"learning_rate": 1.9990555555555557e-06,
"loss": 1.0172,
"step": 72000
},
{
"epoch": 2.93,
"eval_accuracy": 0.49141716566866267,
"eval_loss": 0.9982264041900635,
"eval_runtime": 30.36,
"eval_samples_per_second": 165.02,
"eval_steps_per_second": 20.652,
"step": 72000
},
{
"epoch": 2.97,
"learning_rate": 2.026805555555556e-06,
"loss": 1.0151,
"step": 73000
},
{
"epoch": 2.97,
"eval_accuracy": 0.49720558882235527,
"eval_loss": 0.9918563365936279,
"eval_runtime": 30.3051,
"eval_samples_per_second": 165.318,
"eval_steps_per_second": 20.69,
"step": 73000
},
{
"epoch": 3.01,
"learning_rate": 2.0545833333333335e-06,
"loss": 1.0129,
"step": 74000
},
{
"epoch": 3.01,
"eval_accuracy": 0.49840319361277446,
"eval_loss": 0.9832558631896973,
"eval_runtime": 30.3597,
"eval_samples_per_second": 165.022,
"eval_steps_per_second": 20.652,
"step": 74000
},
{
"epoch": 3.06,
"learning_rate": 2.0823611111111115e-06,
"loss": 1.006,
"step": 75000
},
{
"epoch": 3.06,
"eval_accuracy": 0.5033932135728543,
"eval_loss": 0.9880185127258301,
"eval_runtime": 30.2854,
"eval_samples_per_second": 165.426,
"eval_steps_per_second": 20.703,
"step": 75000
},
{
"epoch": 3.1,
"learning_rate": 2.1101111111111113e-06,
"loss": 1.0069,
"step": 76000
},
{
"epoch": 3.1,
"eval_accuracy": 0.5045908183632735,
"eval_loss": 0.9803372025489807,
"eval_runtime": 30.3285,
"eval_samples_per_second": 165.191,
"eval_steps_per_second": 20.674,
"step": 76000
},
{
"epoch": 3.14,
"learning_rate": 2.137888888888889e-06,
"loss": 1.0106,
"step": 77000
},
{
"epoch": 3.14,
"eval_accuracy": 0.47724550898203594,
"eval_loss": 1.0122599601745605,
"eval_runtime": 30.3162,
"eval_samples_per_second": 165.258,
"eval_steps_per_second": 20.682,
"step": 77000
},
{
"epoch": 3.18,
"learning_rate": 2.1656388888888888e-06,
"loss": 1.0019,
"step": 78000
},
{
"epoch": 3.18,
"eval_accuracy": 0.5045908183632735,
"eval_loss": 0.9800828099250793,
"eval_runtime": 30.2724,
"eval_samples_per_second": 165.497,
"eval_steps_per_second": 20.712,
"step": 78000
},
{
"epoch": 3.22,
"learning_rate": 2.1934166666666667e-06,
"loss": 1.0093,
"step": 79000
},
{
"epoch": 3.22,
"eval_accuracy": 0.493812375249501,
"eval_loss": 0.9990217089653015,
"eval_runtime": 30.2198,
"eval_samples_per_second": 165.785,
"eval_steps_per_second": 20.748,
"step": 79000
},
{
"epoch": 3.26,
"learning_rate": 2.221166666666667e-06,
"loss": 1.0039,
"step": 80000
},
{
"epoch": 3.26,
"eval_accuracy": 0.5067864271457085,
"eval_loss": 0.98616623878479,
"eval_runtime": 30.3059,
"eval_samples_per_second": 165.315,
"eval_steps_per_second": 20.689,
"step": 80000
},
{
"epoch": 3.3,
"learning_rate": 2.2489444444444446e-06,
"loss": 1.0062,
"step": 81000
},
{
"epoch": 3.3,
"eval_accuracy": 0.5013972055888224,
"eval_loss": 0.9929732084274292,
"eval_runtime": 30.2754,
"eval_samples_per_second": 165.481,
"eval_steps_per_second": 20.71,
"step": 81000
},
{
"epoch": 3.34,
"learning_rate": 2.2766944444444444e-06,
"loss": 1.0145,
"step": 82000
},
{
"epoch": 3.34,
"eval_accuracy": 0.5051896207584831,
"eval_loss": 0.9891365766525269,
"eval_runtime": 30.204,
"eval_samples_per_second": 165.872,
"eval_steps_per_second": 20.759,
"step": 82000
},
{
"epoch": 3.38,
"learning_rate": 2.3044722222222224e-06,
"loss": 1.0012,
"step": 83000
},
{
"epoch": 3.38,
"eval_accuracy": 0.5067864271457085,
"eval_loss": 0.984894871711731,
"eval_runtime": 30.2196,
"eval_samples_per_second": 165.787,
"eval_steps_per_second": 20.748,
"step": 83000
},
{
"epoch": 3.42,
"learning_rate": 2.3322222222222223e-06,
"loss": 1.0072,
"step": 84000
},
{
"epoch": 3.42,
"eval_accuracy": 0.5075848303393213,
"eval_loss": 0.979631781578064,
"eval_runtime": 30.1636,
"eval_samples_per_second": 166.094,
"eval_steps_per_second": 20.787,
"step": 84000
},
{
"epoch": 3.46,
"learning_rate": 2.36e-06,
"loss": 1.0038,
"step": 85000
},
{
"epoch": 3.46,
"eval_accuracy": 0.5035928143712575,
"eval_loss": 0.9814818501472473,
"eval_runtime": 30.2922,
"eval_samples_per_second": 165.389,
"eval_steps_per_second": 20.698,
"step": 85000
},
{
"epoch": 3.5,
"learning_rate": 2.38775e-06,
"loss": 1.0107,
"step": 86000
},
{
"epoch": 3.5,
"eval_accuracy": 0.5117764471057884,
"eval_loss": 0.9743403792381287,
"eval_runtime": 30.2926,
"eval_samples_per_second": 165.387,
"eval_steps_per_second": 20.698,
"step": 86000
},
{
"epoch": 3.54,
"learning_rate": 2.415527777777778e-06,
"loss": 1.0011,
"step": 87000
},
{
"epoch": 3.54,
"eval_accuracy": 0.5073852295409181,
"eval_loss": 0.9760627150535583,
"eval_runtime": 30.3925,
"eval_samples_per_second": 164.843,
"eval_steps_per_second": 20.63,
"step": 87000
},
{
"epoch": 3.59,
"learning_rate": 2.443277777777778e-06,
"loss": 1.0045,
"step": 88000
},
{
"epoch": 3.59,
"eval_accuracy": 0.5101796407185629,
"eval_loss": 0.9735654592514038,
"eval_runtime": 30.2088,
"eval_samples_per_second": 165.846,
"eval_steps_per_second": 20.756,
"step": 88000
},
{
"epoch": 3.63,
"learning_rate": 2.4710555555555555e-06,
"loss": 0.999,
"step": 89000
},
{
"epoch": 3.63,
"eval_accuracy": 0.4996007984031936,
"eval_loss": 0.9788551926612854,
"eval_runtime": 30.2584,
"eval_samples_per_second": 165.574,
"eval_steps_per_second": 20.721,
"step": 89000
},
{
"epoch": 3.67,
"learning_rate": 2.498805555555556e-06,
"loss": 1.0075,
"step": 90000
},
{
"epoch": 3.67,
"eval_accuracy": 0.5095808383233533,
"eval_loss": 0.972398579120636,
"eval_runtime": 30.1697,
"eval_samples_per_second": 166.061,
"eval_steps_per_second": 20.782,
"step": 90000
},
{
"epoch": 3.71,
"learning_rate": 2.5265833333333334e-06,
"loss": 1.0085,
"step": 91000
},
{
"epoch": 3.71,
"eval_accuracy": 0.5073852295409181,
"eval_loss": 0.9791931509971619,
"eval_runtime": 30.1687,
"eval_samples_per_second": 166.066,
"eval_steps_per_second": 20.783,
"step": 91000
},
{
"epoch": 3.75,
"learning_rate": 2.5543333333333337e-06,
"loss": 1.0022,
"step": 92000
},
{
"epoch": 3.75,
"eval_accuracy": 0.5105788423153692,
"eval_loss": 0.9835652112960815,
"eval_runtime": 30.389,
"eval_samples_per_second": 164.862,
"eval_steps_per_second": 20.632,
"step": 92000
},
{
"epoch": 3.79,
"learning_rate": 2.5821111111111112e-06,
"loss": 1.006,
"step": 93000
},
{
"epoch": 3.79,
"eval_accuracy": 0.5127744510978044,
"eval_loss": 0.9716958403587341,
"eval_runtime": 30.2822,
"eval_samples_per_second": 165.444,
"eval_steps_per_second": 20.705,
"step": 93000
},
{
"epoch": 3.83,
"learning_rate": 2.609861111111111e-06,
"loss": 1.0001,
"step": 94000
},
{
"epoch": 3.83,
"eval_accuracy": 0.49880239520958086,
"eval_loss": 0.982902467250824,
"eval_runtime": 30.3141,
"eval_samples_per_second": 165.27,
"eval_steps_per_second": 20.683,
"step": 94000
},
{
"epoch": 3.87,
"learning_rate": 2.637638888888889e-06,
"loss": 1.0018,
"step": 95000
},
{
"epoch": 3.87,
"eval_accuracy": 0.5117764471057884,
"eval_loss": 0.9711886644363403,
"eval_runtime": 30.3419,
"eval_samples_per_second": 165.118,
"eval_steps_per_second": 20.665,
"step": 95000
},
{
"epoch": 3.91,
"learning_rate": 2.6653888888888894e-06,
"loss": 1.001,
"step": 96000
},
{
"epoch": 3.91,
"eval_accuracy": 0.5085828343313373,
"eval_loss": 0.9692357182502747,
"eval_runtime": 30.2793,
"eval_samples_per_second": 165.46,
"eval_steps_per_second": 20.707,
"step": 96000
},
{
"epoch": 3.95,
"learning_rate": 2.693166666666667e-06,
"loss": 0.9967,
"step": 97000
},
{
"epoch": 3.95,
"eval_accuracy": 0.5035928143712575,
"eval_loss": 0.9811190366744995,
"eval_runtime": 30.3522,
"eval_samples_per_second": 165.062,
"eval_steps_per_second": 20.657,
"step": 97000
},
{
"epoch": 3.99,
"learning_rate": 2.7209166666666668e-06,
"loss": 0.9963,
"step": 98000
},
{
"epoch": 3.99,
"eval_accuracy": 0.5117764471057884,
"eval_loss": 0.9744167923927307,
"eval_runtime": 30.2586,
"eval_samples_per_second": 165.573,
"eval_steps_per_second": 20.721,
"step": 98000
},
{
"epoch": 4.03,
"learning_rate": 2.7486944444444448e-06,
"loss": 0.9923,
"step": 99000
},
{
"epoch": 4.03,
"eval_accuracy": 0.5155688622754491,
"eval_loss": 0.9705401659011841,
"eval_runtime": 30.3652,
"eval_samples_per_second": 164.992,
"eval_steps_per_second": 20.649,
"step": 99000
},
{
"epoch": 4.07,
"learning_rate": 2.7764444444444446e-06,
"loss": 0.9925,
"step": 100000
},
{
"epoch": 4.07,
"eval_accuracy": 0.5161676646706587,
"eval_loss": 0.967511773109436,
"eval_runtime": 30.3989,
"eval_samples_per_second": 164.809,
"eval_steps_per_second": 20.626,
"step": 100000
},
{
"epoch": 4.12,
"learning_rate": 2.804222222222222e-06,
"loss": 0.9965,
"step": 101000
},
{
"epoch": 4.12,
"eval_accuracy": 0.5179640718562875,
"eval_loss": 0.9703009724617004,
"eval_runtime": 30.306,
"eval_samples_per_second": 165.314,
"eval_steps_per_second": 20.689,
"step": 101000
},
{
"epoch": 4.16,
"learning_rate": 2.8319722222222225e-06,
"loss": 0.994,
"step": 102000
},
{
"epoch": 4.16,
"eval_accuracy": 0.5027944111776447,
"eval_loss": 0.9799131155014038,
"eval_runtime": 30.4057,
"eval_samples_per_second": 164.772,
"eval_steps_per_second": 20.621,
"step": 102000
},
{
"epoch": 4.2,
"learning_rate": 2.8597500000000004e-06,
"loss": 0.9989,
"step": 103000
},
{
"epoch": 4.2,
"eval_accuracy": 0.5175648702594811,
"eval_loss": 0.965115487575531,
"eval_runtime": 30.5041,
"eval_samples_per_second": 164.24,
"eval_steps_per_second": 20.555,
"step": 103000
},
{
"epoch": 4.24,
"learning_rate": 2.8875000000000003e-06,
"loss": 0.9909,
"step": 104000
},
{
"epoch": 4.24,
"eval_accuracy": 0.5005988023952096,
"eval_loss": 0.9864978790283203,
"eval_runtime": 30.393,
"eval_samples_per_second": 164.84,
"eval_steps_per_second": 20.63,
"step": 104000
},
{
"epoch": 4.28,
"learning_rate": 2.915277777777778e-06,
"loss": 0.9873,
"step": 105000
},
{
"epoch": 4.28,
"eval_accuracy": 0.5017964071856288,
"eval_loss": 0.9913762211799622,
"eval_runtime": 30.3351,
"eval_samples_per_second": 165.155,
"eval_steps_per_second": 20.669,
"step": 105000
},
{
"epoch": 4.32,
"learning_rate": 2.9430277777777777e-06,
"loss": 0.9826,
"step": 106000
},
{
"epoch": 4.32,
"eval_accuracy": 0.5201596806387225,
"eval_loss": 0.9676294922828674,
"eval_runtime": 30.2604,
"eval_samples_per_second": 165.563,
"eval_steps_per_second": 20.72,
"step": 106000
},
{
"epoch": 4.36,
"learning_rate": 2.9708055555555557e-06,
"loss": 0.9832,
"step": 107000
},
{
"epoch": 4.36,
"eval_accuracy": 0.5173652694610779,
"eval_loss": 0.9818444848060608,
"eval_runtime": 30.3828,
"eval_samples_per_second": 164.896,
"eval_steps_per_second": 20.637,
"step": 107000
},
{
"epoch": 4.4,
"learning_rate": 2.998555555555556e-06,
"loss": 0.984,
"step": 108000
},
{
"epoch": 4.4,
"eval_accuracy": 0.511377245508982,
"eval_loss": 0.9720831513404846,
"eval_runtime": 30.2651,
"eval_samples_per_second": 165.537,
"eval_steps_per_second": 20.717,
"step": 108000
},
{
"epoch": 4.44,
"learning_rate": 3.0263333333333336e-06,
"loss": 0.982,
"step": 109000
},
{
"epoch": 4.44,
"eval_accuracy": 0.5209580838323353,
"eval_loss": 0.9712573885917664,
"eval_runtime": 30.204,
"eval_samples_per_second": 165.872,
"eval_steps_per_second": 20.759,
"step": 109000
},
{
"epoch": 4.48,
"learning_rate": 3.0540833333333334e-06,
"loss": 0.9876,
"step": 110000
},
{
"epoch": 4.48,
"eval_accuracy": 0.5265469061876248,
"eval_loss": 0.9630448222160339,
"eval_runtime": 30.2241,
"eval_samples_per_second": 165.762,
"eval_steps_per_second": 20.745,
"step": 110000
},
{
"epoch": 4.52,
"learning_rate": 3.0818611111111114e-06,
"loss": 0.9826,
"step": 111000
},
{
"epoch": 4.52,
"eval_accuracy": 0.524750499001996,
"eval_loss": 0.9607908129692078,
"eval_runtime": 30.5687,
"eval_samples_per_second": 163.893,
"eval_steps_per_second": 20.511,
"step": 111000
}
],
"logging_steps": 1000,
"max_steps": 10000000,
"num_train_epochs": 408,
"save_steps": 1000,
"total_flos": 1.1601732971475763e+17,
"trial_name": null,
"trial_params": null
}