yt-special-batch12-small / trainer_state.json
TheRains's picture
End of training
29f7a2b
{
"best_metric": 40.08170676350431,
"best_model_checkpoint": "./output/small/yt-special-batch12-small/checkpoint-4000",
"epoch": 1.2883277505797475,
"global_step": 5000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 4.2000000000000006e-07,
"loss": 2.3125,
"step": 25
},
{
"epoch": 0.01,
"learning_rate": 9.200000000000001e-07,
"loss": 2.2663,
"step": 50
},
{
"epoch": 0.02,
"learning_rate": 1.42e-06,
"loss": 1.7932,
"step": 75
},
{
"epoch": 0.03,
"learning_rate": 1.9200000000000003e-06,
"loss": 1.5653,
"step": 100
},
{
"epoch": 0.03,
"learning_rate": 2.42e-06,
"loss": 1.1108,
"step": 125
},
{
"epoch": 0.04,
"learning_rate": 2.92e-06,
"loss": 1.0929,
"step": 150
},
{
"epoch": 0.05,
"learning_rate": 3.4200000000000007e-06,
"loss": 0.9807,
"step": 175
},
{
"epoch": 0.05,
"learning_rate": 3.920000000000001e-06,
"loss": 1.0657,
"step": 200
},
{
"epoch": 0.06,
"learning_rate": 4.42e-06,
"loss": 0.9355,
"step": 225
},
{
"epoch": 0.06,
"learning_rate": 4.92e-06,
"loss": 1.016,
"step": 250
},
{
"epoch": 0.07,
"learning_rate": 5.420000000000001e-06,
"loss": 0.8875,
"step": 275
},
{
"epoch": 0.08,
"learning_rate": 5.92e-06,
"loss": 0.9298,
"step": 300
},
{
"epoch": 0.08,
"learning_rate": 6.42e-06,
"loss": 0.9081,
"step": 325
},
{
"epoch": 0.09,
"learning_rate": 6.92e-06,
"loss": 0.9712,
"step": 350
},
{
"epoch": 0.1,
"learning_rate": 7.420000000000001e-06,
"loss": 0.9151,
"step": 375
},
{
"epoch": 0.1,
"learning_rate": 7.92e-06,
"loss": 0.9823,
"step": 400
},
{
"epoch": 0.11,
"learning_rate": 8.42e-06,
"loss": 0.9278,
"step": 425
},
{
"epoch": 0.12,
"learning_rate": 8.920000000000001e-06,
"loss": 0.9049,
"step": 450
},
{
"epoch": 0.12,
"learning_rate": 9.42e-06,
"loss": 0.8354,
"step": 475
},
{
"epoch": 0.13,
"learning_rate": 9.920000000000002e-06,
"loss": 0.8975,
"step": 500
},
{
"epoch": 0.14,
"learning_rate": 9.953333333333333e-06,
"loss": 0.8576,
"step": 525
},
{
"epoch": 0.14,
"learning_rate": 9.89777777777778e-06,
"loss": 0.81,
"step": 550
},
{
"epoch": 0.15,
"learning_rate": 9.842222222222223e-06,
"loss": 0.8336,
"step": 575
},
{
"epoch": 0.15,
"learning_rate": 9.786666666666667e-06,
"loss": 0.8869,
"step": 600
},
{
"epoch": 0.16,
"learning_rate": 9.731111111111113e-06,
"loss": 0.8731,
"step": 625
},
{
"epoch": 0.17,
"learning_rate": 9.675555555555555e-06,
"loss": 0.8089,
"step": 650
},
{
"epoch": 0.17,
"learning_rate": 9.620000000000001e-06,
"loss": 0.8427,
"step": 675
},
{
"epoch": 0.18,
"learning_rate": 9.564444444444445e-06,
"loss": 0.8825,
"step": 700
},
{
"epoch": 0.19,
"learning_rate": 9.508888888888889e-06,
"loss": 0.901,
"step": 725
},
{
"epoch": 0.19,
"learning_rate": 9.453333333333335e-06,
"loss": 0.8191,
"step": 750
},
{
"epoch": 0.2,
"learning_rate": 9.397777777777779e-06,
"loss": 0.9662,
"step": 775
},
{
"epoch": 0.21,
"learning_rate": 9.342222222222223e-06,
"loss": 0.8861,
"step": 800
},
{
"epoch": 0.21,
"learning_rate": 9.286666666666667e-06,
"loss": 0.8501,
"step": 825
},
{
"epoch": 0.22,
"learning_rate": 9.231111111111111e-06,
"loss": 0.7936,
"step": 850
},
{
"epoch": 0.23,
"learning_rate": 9.175555555555557e-06,
"loss": 0.8678,
"step": 875
},
{
"epoch": 0.23,
"learning_rate": 9.12e-06,
"loss": 0.7968,
"step": 900
},
{
"epoch": 0.24,
"learning_rate": 9.064444444444447e-06,
"loss": 0.8059,
"step": 925
},
{
"epoch": 0.24,
"learning_rate": 9.008888888888889e-06,
"loss": 0.8562,
"step": 950
},
{
"epoch": 0.25,
"learning_rate": 8.953333333333335e-06,
"loss": 0.8176,
"step": 975
},
{
"epoch": 0.26,
"learning_rate": 8.897777777777779e-06,
"loss": 0.8104,
"step": 1000
},
{
"epoch": 0.26,
"eval_loss": 0.8243614435195923,
"eval_runtime": 1697.0003,
"eval_samples_per_second": 1.444,
"eval_steps_per_second": 0.241,
"eval_wer": 49.73737111730757,
"step": 1000
},
{
"epoch": 0.26,
"learning_rate": 8.842222222222223e-06,
"loss": 0.8632,
"step": 1025
},
{
"epoch": 0.27,
"learning_rate": 8.786666666666668e-06,
"loss": 0.8458,
"step": 1050
},
{
"epoch": 0.28,
"learning_rate": 8.73111111111111e-06,
"loss": 0.7638,
"step": 1075
},
{
"epoch": 0.28,
"learning_rate": 8.675555555555556e-06,
"loss": 0.8451,
"step": 1100
},
{
"epoch": 0.29,
"learning_rate": 8.62e-06,
"loss": 0.8284,
"step": 1125
},
{
"epoch": 0.3,
"learning_rate": 8.564444444444445e-06,
"loss": 0.7474,
"step": 1150
},
{
"epoch": 0.3,
"learning_rate": 8.50888888888889e-06,
"loss": 0.8102,
"step": 1175
},
{
"epoch": 0.31,
"learning_rate": 8.453333333333334e-06,
"loss": 0.7658,
"step": 1200
},
{
"epoch": 0.32,
"learning_rate": 8.397777777777778e-06,
"loss": 0.8303,
"step": 1225
},
{
"epoch": 0.32,
"learning_rate": 8.342222222222222e-06,
"loss": 0.7827,
"step": 1250
},
{
"epoch": 0.33,
"learning_rate": 8.286666666666668e-06,
"loss": 0.8155,
"step": 1275
},
{
"epoch": 0.33,
"learning_rate": 8.231111111111112e-06,
"loss": 0.8411,
"step": 1300
},
{
"epoch": 0.34,
"learning_rate": 8.175555555555556e-06,
"loss": 0.7805,
"step": 1325
},
{
"epoch": 0.35,
"learning_rate": 8.120000000000002e-06,
"loss": 0.774,
"step": 1350
},
{
"epoch": 0.35,
"learning_rate": 8.064444444444444e-06,
"loss": 0.8251,
"step": 1375
},
{
"epoch": 0.36,
"learning_rate": 8.00888888888889e-06,
"loss": 0.7772,
"step": 1400
},
{
"epoch": 0.37,
"learning_rate": 7.953333333333334e-06,
"loss": 0.7266,
"step": 1425
},
{
"epoch": 0.37,
"learning_rate": 7.897777777777778e-06,
"loss": 0.833,
"step": 1450
},
{
"epoch": 0.38,
"learning_rate": 7.842222222222224e-06,
"loss": 0.8107,
"step": 1475
},
{
"epoch": 0.39,
"learning_rate": 7.786666666666666e-06,
"loss": 0.7071,
"step": 1500
},
{
"epoch": 0.39,
"learning_rate": 7.731111111111112e-06,
"loss": 0.7231,
"step": 1525
},
{
"epoch": 0.4,
"learning_rate": 7.675555555555556e-06,
"loss": 0.8016,
"step": 1550
},
{
"epoch": 0.41,
"learning_rate": 7.620000000000001e-06,
"loss": 0.8692,
"step": 1575
},
{
"epoch": 0.41,
"learning_rate": 7.564444444444446e-06,
"loss": 0.7495,
"step": 1600
},
{
"epoch": 0.42,
"learning_rate": 7.50888888888889e-06,
"loss": 0.7892,
"step": 1625
},
{
"epoch": 0.43,
"learning_rate": 7.453333333333334e-06,
"loss": 0.7232,
"step": 1650
},
{
"epoch": 0.43,
"learning_rate": 7.3977777777777786e-06,
"loss": 0.7565,
"step": 1675
},
{
"epoch": 0.44,
"learning_rate": 7.342222222222223e-06,
"loss": 0.7897,
"step": 1700
},
{
"epoch": 0.44,
"learning_rate": 7.2866666666666675e-06,
"loss": 0.7688,
"step": 1725
},
{
"epoch": 0.45,
"learning_rate": 7.231111111111112e-06,
"loss": 0.7258,
"step": 1750
},
{
"epoch": 0.46,
"learning_rate": 7.1755555555555556e-06,
"loss": 0.7703,
"step": 1775
},
{
"epoch": 0.46,
"learning_rate": 7.1200000000000004e-06,
"loss": 0.8285,
"step": 1800
},
{
"epoch": 0.47,
"learning_rate": 7.0644444444444445e-06,
"loss": 0.7844,
"step": 1825
},
{
"epoch": 0.48,
"learning_rate": 7.008888888888889e-06,
"loss": 0.8188,
"step": 1850
},
{
"epoch": 0.48,
"learning_rate": 6.953333333333334e-06,
"loss": 0.6785,
"step": 1875
},
{
"epoch": 0.49,
"learning_rate": 6.897777777777779e-06,
"loss": 0.743,
"step": 1900
},
{
"epoch": 0.5,
"learning_rate": 6.842222222222222e-06,
"loss": 0.7725,
"step": 1925
},
{
"epoch": 0.5,
"learning_rate": 6.786666666666667e-06,
"loss": 0.7709,
"step": 1950
},
{
"epoch": 0.51,
"learning_rate": 6.731111111111111e-06,
"loss": 0.8199,
"step": 1975
},
{
"epoch": 0.52,
"learning_rate": 6.675555555555556e-06,
"loss": 0.7059,
"step": 2000
},
{
"epoch": 0.52,
"eval_loss": 0.7380304336547852,
"eval_runtime": 1732.1704,
"eval_samples_per_second": 1.415,
"eval_steps_per_second": 0.236,
"eval_wer": 47.967057908047465,
"step": 2000
},
{
"epoch": 0.52,
"learning_rate": 6.620000000000001e-06,
"loss": 0.7308,
"step": 2025
},
{
"epoch": 0.53,
"learning_rate": 6.564444444444446e-06,
"loss": 0.6404,
"step": 2050
},
{
"epoch": 0.53,
"learning_rate": 6.508888888888889e-06,
"loss": 0.7401,
"step": 2075
},
{
"epoch": 0.54,
"learning_rate": 6.453333333333334e-06,
"loss": 0.7784,
"step": 2100
},
{
"epoch": 0.55,
"learning_rate": 6.397777777777778e-06,
"loss": 0.7662,
"step": 2125
},
{
"epoch": 0.55,
"learning_rate": 6.342222222222223e-06,
"loss": 0.7704,
"step": 2150
},
{
"epoch": 0.56,
"learning_rate": 6.286666666666668e-06,
"loss": 0.7226,
"step": 2175
},
{
"epoch": 0.57,
"learning_rate": 6.231111111111111e-06,
"loss": 0.6753,
"step": 2200
},
{
"epoch": 0.57,
"learning_rate": 6.175555555555556e-06,
"loss": 0.6995,
"step": 2225
},
{
"epoch": 0.58,
"learning_rate": 6.120000000000001e-06,
"loss": 0.6929,
"step": 2250
},
{
"epoch": 0.59,
"learning_rate": 6.064444444444445e-06,
"loss": 0.699,
"step": 2275
},
{
"epoch": 0.59,
"learning_rate": 6.00888888888889e-06,
"loss": 0.8008,
"step": 2300
},
{
"epoch": 0.6,
"learning_rate": 5.9533333333333345e-06,
"loss": 0.8015,
"step": 2325
},
{
"epoch": 0.61,
"learning_rate": 5.897777777777778e-06,
"loss": 0.7418,
"step": 2350
},
{
"epoch": 0.61,
"learning_rate": 5.8422222222222226e-06,
"loss": 0.673,
"step": 2375
},
{
"epoch": 0.62,
"learning_rate": 5.7866666666666674e-06,
"loss": 0.689,
"step": 2400
},
{
"epoch": 0.62,
"learning_rate": 5.7311111111111115e-06,
"loss": 0.731,
"step": 2425
},
{
"epoch": 0.63,
"learning_rate": 5.675555555555556e-06,
"loss": 0.7157,
"step": 2450
},
{
"epoch": 0.64,
"learning_rate": 5.620000000000001e-06,
"loss": 0.7544,
"step": 2475
},
{
"epoch": 0.64,
"learning_rate": 5.5644444444444444e-06,
"loss": 0.6883,
"step": 2500
},
{
"epoch": 0.65,
"learning_rate": 5.508888888888889e-06,
"loss": 0.6996,
"step": 2525
},
{
"epoch": 0.66,
"learning_rate": 5.453333333333334e-06,
"loss": 0.6992,
"step": 2550
},
{
"epoch": 0.66,
"learning_rate": 5.397777777777778e-06,
"loss": 0.7111,
"step": 2575
},
{
"epoch": 0.67,
"learning_rate": 5.342222222222223e-06,
"loss": 0.7444,
"step": 2600
},
{
"epoch": 0.68,
"learning_rate": 5.286666666666666e-06,
"loss": 0.6413,
"step": 2625
},
{
"epoch": 0.68,
"learning_rate": 5.231111111111111e-06,
"loss": 0.6875,
"step": 2650
},
{
"epoch": 0.69,
"learning_rate": 5.175555555555556e-06,
"loss": 0.7132,
"step": 2675
},
{
"epoch": 0.7,
"learning_rate": 5.12e-06,
"loss": 0.7571,
"step": 2700
},
{
"epoch": 0.7,
"learning_rate": 5.064444444444445e-06,
"loss": 0.6727,
"step": 2725
},
{
"epoch": 0.71,
"learning_rate": 5.00888888888889e-06,
"loss": 0.6749,
"step": 2750
},
{
"epoch": 0.72,
"learning_rate": 4.953333333333334e-06,
"loss": 0.7157,
"step": 2775
},
{
"epoch": 0.72,
"learning_rate": 4.897777777777778e-06,
"loss": 0.7161,
"step": 2800
},
{
"epoch": 0.73,
"learning_rate": 4.842222222222223e-06,
"loss": 0.7021,
"step": 2825
},
{
"epoch": 0.73,
"learning_rate": 4.786666666666667e-06,
"loss": 0.7914,
"step": 2850
},
{
"epoch": 0.74,
"learning_rate": 4.731111111111112e-06,
"loss": 0.6452,
"step": 2875
},
{
"epoch": 0.75,
"learning_rate": 4.675555555555556e-06,
"loss": 0.7202,
"step": 2900
},
{
"epoch": 0.75,
"learning_rate": 4.620000000000001e-06,
"loss": 0.7219,
"step": 2925
},
{
"epoch": 0.76,
"learning_rate": 4.564444444444445e-06,
"loss": 0.7599,
"step": 2950
},
{
"epoch": 0.77,
"learning_rate": 4.50888888888889e-06,
"loss": 0.7362,
"step": 2975
},
{
"epoch": 0.77,
"learning_rate": 4.453333333333334e-06,
"loss": 0.7127,
"step": 3000
},
{
"epoch": 0.77,
"eval_loss": 0.6957079172134399,
"eval_runtime": 1754.3478,
"eval_samples_per_second": 1.397,
"eval_steps_per_second": 0.233,
"eval_wer": 48.83600285325206,
"step": 3000
},
{
"epoch": 0.78,
"learning_rate": 4.397777777777778e-06,
"loss": 0.6489,
"step": 3025
},
{
"epoch": 0.79,
"learning_rate": 4.3422222222222225e-06,
"loss": 0.7291,
"step": 3050
},
{
"epoch": 0.79,
"learning_rate": 4.2866666666666666e-06,
"loss": 0.67,
"step": 3075
},
{
"epoch": 0.8,
"learning_rate": 4.2311111111111114e-06,
"loss": 0.7401,
"step": 3100
},
{
"epoch": 0.81,
"learning_rate": 4.175555555555556e-06,
"loss": 0.7096,
"step": 3125
},
{
"epoch": 0.81,
"learning_rate": 4.12e-06,
"loss": 0.7444,
"step": 3150
},
{
"epoch": 0.82,
"learning_rate": 4.064444444444444e-06,
"loss": 0.611,
"step": 3175
},
{
"epoch": 0.82,
"learning_rate": 4.008888888888889e-06,
"loss": 0.7566,
"step": 3200
},
{
"epoch": 0.83,
"learning_rate": 3.953333333333333e-06,
"loss": 0.662,
"step": 3225
},
{
"epoch": 0.84,
"learning_rate": 3.897777777777778e-06,
"loss": 0.6092,
"step": 3250
},
{
"epoch": 0.84,
"learning_rate": 3.842222222222223e-06,
"loss": 0.6924,
"step": 3275
},
{
"epoch": 0.85,
"learning_rate": 3.7866666666666667e-06,
"loss": 0.6527,
"step": 3300
},
{
"epoch": 0.86,
"learning_rate": 3.7311111111111116e-06,
"loss": 0.5968,
"step": 3325
},
{
"epoch": 0.86,
"learning_rate": 3.675555555555556e-06,
"loss": 0.6793,
"step": 3350
},
{
"epoch": 0.87,
"learning_rate": 3.62e-06,
"loss": 0.6826,
"step": 3375
},
{
"epoch": 0.88,
"learning_rate": 3.564444444444445e-06,
"loss": 0.7087,
"step": 3400
},
{
"epoch": 0.88,
"learning_rate": 3.508888888888889e-06,
"loss": 0.6763,
"step": 3425
},
{
"epoch": 0.89,
"learning_rate": 3.4533333333333334e-06,
"loss": 0.6555,
"step": 3450
},
{
"epoch": 0.9,
"learning_rate": 3.3977777777777783e-06,
"loss": 0.6338,
"step": 3475
},
{
"epoch": 0.9,
"learning_rate": 3.3422222222222224e-06,
"loss": 0.6586,
"step": 3500
},
{
"epoch": 0.91,
"learning_rate": 3.286666666666667e-06,
"loss": 0.6893,
"step": 3525
},
{
"epoch": 0.91,
"learning_rate": 3.2311111111111117e-06,
"loss": 0.6377,
"step": 3550
},
{
"epoch": 0.92,
"learning_rate": 3.1755555555555557e-06,
"loss": 0.6808,
"step": 3575
},
{
"epoch": 0.93,
"learning_rate": 3.12e-06,
"loss": 0.6608,
"step": 3600
},
{
"epoch": 0.93,
"learning_rate": 3.064444444444445e-06,
"loss": 0.7201,
"step": 3625
},
{
"epoch": 0.94,
"learning_rate": 3.008888888888889e-06,
"loss": 0.6794,
"step": 3650
},
{
"epoch": 0.95,
"learning_rate": 2.9533333333333336e-06,
"loss": 0.6893,
"step": 3675
},
{
"epoch": 0.95,
"learning_rate": 2.8977777777777785e-06,
"loss": 0.5359,
"step": 3700
},
{
"epoch": 0.96,
"learning_rate": 2.8422222222222225e-06,
"loss": 0.645,
"step": 3725
},
{
"epoch": 0.97,
"learning_rate": 2.786666666666667e-06,
"loss": 0.693,
"step": 3750
},
{
"epoch": 0.97,
"learning_rate": 2.7311111111111114e-06,
"loss": 0.6164,
"step": 3775
},
{
"epoch": 0.98,
"learning_rate": 2.675555555555556e-06,
"loss": 0.611,
"step": 3800
},
{
"epoch": 0.99,
"learning_rate": 2.6200000000000003e-06,
"loss": 0.7129,
"step": 3825
},
{
"epoch": 0.99,
"learning_rate": 2.5644444444444444e-06,
"loss": 0.6676,
"step": 3850
},
{
"epoch": 1.0,
"learning_rate": 2.5088888888888892e-06,
"loss": 0.706,
"step": 3875
},
{
"epoch": 1.0,
"learning_rate": 2.4533333333333333e-06,
"loss": 0.5449,
"step": 3900
},
{
"epoch": 1.01,
"learning_rate": 2.397777777777778e-06,
"loss": 0.4825,
"step": 3925
},
{
"epoch": 1.02,
"learning_rate": 2.342222222222222e-06,
"loss": 0.3935,
"step": 3950
},
{
"epoch": 1.02,
"learning_rate": 2.2866666666666667e-06,
"loss": 0.461,
"step": 3975
},
{
"epoch": 1.03,
"learning_rate": 2.2311111111111115e-06,
"loss": 0.5311,
"step": 4000
},
{
"epoch": 1.03,
"eval_loss": 0.6717672944068909,
"eval_runtime": 1672.3774,
"eval_samples_per_second": 1.466,
"eval_steps_per_second": 0.245,
"eval_wer": 40.08170676350431,
"step": 4000
},
{
"epoch": 1.04,
"learning_rate": 2.1755555555555556e-06,
"loss": 0.4439,
"step": 4025
},
{
"epoch": 1.04,
"learning_rate": 2.12e-06,
"loss": 0.4455,
"step": 4050
},
{
"epoch": 1.05,
"learning_rate": 2.064444444444445e-06,
"loss": 0.4387,
"step": 4075
},
{
"epoch": 1.06,
"learning_rate": 2.008888888888889e-06,
"loss": 0.4408,
"step": 4100
},
{
"epoch": 1.06,
"learning_rate": 1.9533333333333334e-06,
"loss": 0.497,
"step": 4125
},
{
"epoch": 1.07,
"learning_rate": 1.8977777777777779e-06,
"loss": 0.5517,
"step": 4150
},
{
"epoch": 1.08,
"learning_rate": 1.8422222222222225e-06,
"loss": 0.4741,
"step": 4175
},
{
"epoch": 1.08,
"learning_rate": 1.7866666666666668e-06,
"loss": 0.4495,
"step": 4200
},
{
"epoch": 1.09,
"learning_rate": 1.7311111111111112e-06,
"loss": 0.4527,
"step": 4225
},
{
"epoch": 1.1,
"learning_rate": 1.675555555555556e-06,
"loss": 0.4564,
"step": 4250
},
{
"epoch": 1.1,
"learning_rate": 1.6200000000000002e-06,
"loss": 0.4253,
"step": 4275
},
{
"epoch": 1.11,
"learning_rate": 1.5644444444444446e-06,
"loss": 0.4263,
"step": 4300
},
{
"epoch": 1.11,
"learning_rate": 1.5088888888888889e-06,
"loss": 0.486,
"step": 4325
},
{
"epoch": 1.12,
"learning_rate": 1.4533333333333335e-06,
"loss": 0.4233,
"step": 4350
},
{
"epoch": 1.13,
"learning_rate": 1.397777777777778e-06,
"loss": 0.4304,
"step": 4375
},
{
"epoch": 1.13,
"learning_rate": 1.3422222222222222e-06,
"loss": 0.4511,
"step": 4400
},
{
"epoch": 1.14,
"learning_rate": 1.286666666666667e-06,
"loss": 0.4845,
"step": 4425
},
{
"epoch": 1.15,
"learning_rate": 1.2311111111111112e-06,
"loss": 0.4074,
"step": 4450
},
{
"epoch": 1.15,
"learning_rate": 1.1755555555555556e-06,
"loss": 0.4087,
"step": 4475
},
{
"epoch": 1.16,
"learning_rate": 1.12e-06,
"loss": 0.3898,
"step": 4500
},
{
"epoch": 1.17,
"learning_rate": 1.0644444444444445e-06,
"loss": 0.4003,
"step": 4525
},
{
"epoch": 1.17,
"learning_rate": 1.008888888888889e-06,
"loss": 0.4805,
"step": 4550
},
{
"epoch": 1.18,
"learning_rate": 9.533333333333335e-07,
"loss": 0.4217,
"step": 4575
},
{
"epoch": 1.19,
"learning_rate": 8.977777777777778e-07,
"loss": 0.4657,
"step": 4600
},
{
"epoch": 1.19,
"learning_rate": 8.422222222222224e-07,
"loss": 0.4665,
"step": 4625
},
{
"epoch": 1.2,
"learning_rate": 7.866666666666667e-07,
"loss": 0.4675,
"step": 4650
},
{
"epoch": 1.2,
"learning_rate": 7.311111111111112e-07,
"loss": 0.4278,
"step": 4675
},
{
"epoch": 1.21,
"learning_rate": 6.755555555555555e-07,
"loss": 0.4513,
"step": 4700
},
{
"epoch": 1.22,
"learning_rate": 6.200000000000001e-07,
"loss": 0.5258,
"step": 4725
},
{
"epoch": 1.22,
"learning_rate": 5.644444444444445e-07,
"loss": 0.4163,
"step": 4750
},
{
"epoch": 1.23,
"learning_rate": 5.088888888888889e-07,
"loss": 0.4538,
"step": 4775
},
{
"epoch": 1.24,
"learning_rate": 4.533333333333334e-07,
"loss": 0.4305,
"step": 4800
},
{
"epoch": 1.24,
"learning_rate": 3.9777777777777783e-07,
"loss": 0.4332,
"step": 4825
},
{
"epoch": 1.25,
"learning_rate": 3.422222222222223e-07,
"loss": 0.4289,
"step": 4850
},
{
"epoch": 1.26,
"learning_rate": 2.866666666666667e-07,
"loss": 0.4568,
"step": 4875
},
{
"epoch": 1.26,
"learning_rate": 2.3111111111111112e-07,
"loss": 0.452,
"step": 4900
},
{
"epoch": 1.27,
"learning_rate": 1.7555555555555558e-07,
"loss": 0.4661,
"step": 4925
},
{
"epoch": 1.28,
"learning_rate": 1.2000000000000002e-07,
"loss": 0.4143,
"step": 4950
},
{
"epoch": 1.28,
"learning_rate": 6.444444444444445e-08,
"loss": 0.3722,
"step": 4975
},
{
"epoch": 1.29,
"learning_rate": 8.88888888888889e-09,
"loss": 0.47,
"step": 5000
},
{
"epoch": 1.29,
"eval_loss": 0.6645243763923645,
"eval_runtime": 1666.9725,
"eval_samples_per_second": 1.47,
"eval_steps_per_second": 0.245,
"eval_wer": 40.42539394332404,
"step": 5000
},
{
"epoch": 1.29,
"step": 5000,
"total_flos": 1.731454684839936e+19,
"train_loss": 0.7206048248291016,
"train_runtime": 45959.6141,
"train_samples_per_second": 1.305,
"train_steps_per_second": 0.109
}
],
"max_steps": 5000,
"num_train_epochs": 2,
"total_flos": 1.731454684839936e+19,
"trial_name": null,
"trial_params": null
}