yt-special-batch8-lr6-small / trainer_state.json
TheRains's picture
End of training
75e8938
{
"best_metric": 53.80973996498282,
"best_model_checkpoint": "./output/small/yt-special-batch8-lr6-small/checkpoint-4000",
"epoch": 0.8588114050154586,
"global_step": 5000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 4.2e-08,
"loss": 2.3663,
"step": 25
},
{
"epoch": 0.01,
"learning_rate": 9.199999999999999e-08,
"loss": 2.2617,
"step": 50
},
{
"epoch": 0.01,
"learning_rate": 1.4199999999999997e-07,
"loss": 2.4728,
"step": 75
},
{
"epoch": 0.02,
"learning_rate": 1.92e-07,
"loss": 2.1718,
"step": 100
},
{
"epoch": 0.02,
"learning_rate": 2.4e-07,
"loss": 2.1848,
"step": 125
},
{
"epoch": 0.03,
"learning_rate": 2.9e-07,
"loss": 2.156,
"step": 150
},
{
"epoch": 0.03,
"learning_rate": 3.4000000000000003e-07,
"loss": 1.8236,
"step": 175
},
{
"epoch": 0.03,
"learning_rate": 3.8999999999999997e-07,
"loss": 1.8415,
"step": 200
},
{
"epoch": 0.04,
"learning_rate": 4.3999999999999997e-07,
"loss": 1.8769,
"step": 225
},
{
"epoch": 0.04,
"learning_rate": 4.9e-07,
"loss": 1.7158,
"step": 250
},
{
"epoch": 0.05,
"learning_rate": 5.4e-07,
"loss": 1.6218,
"step": 275
},
{
"epoch": 0.05,
"learning_rate": 5.9e-07,
"loss": 1.5328,
"step": 300
},
{
"epoch": 0.06,
"learning_rate": 6.4e-07,
"loss": 1.2934,
"step": 325
},
{
"epoch": 0.06,
"learning_rate": 6.9e-07,
"loss": 1.3013,
"step": 350
},
{
"epoch": 0.06,
"learning_rate": 7.4e-07,
"loss": 1.295,
"step": 375
},
{
"epoch": 0.07,
"learning_rate": 7.9e-07,
"loss": 1.1418,
"step": 400
},
{
"epoch": 0.07,
"learning_rate": 8.399999999999999e-07,
"loss": 1.0011,
"step": 425
},
{
"epoch": 0.08,
"learning_rate": 8.9e-07,
"loss": 1.1557,
"step": 450
},
{
"epoch": 0.08,
"learning_rate": 9.399999999999999e-07,
"loss": 1.107,
"step": 475
},
{
"epoch": 0.09,
"learning_rate": 9.9e-07,
"loss": 1.1109,
"step": 500
},
{
"epoch": 0.09,
"learning_rate": 9.955555555555554e-07,
"loss": 1.042,
"step": 525
},
{
"epoch": 0.09,
"learning_rate": 9.9e-07,
"loss": 1.0429,
"step": 550
},
{
"epoch": 0.1,
"learning_rate": 9.844444444444445e-07,
"loss": 0.9905,
"step": 575
},
{
"epoch": 0.1,
"learning_rate": 9.788888888888889e-07,
"loss": 1.1513,
"step": 600
},
{
"epoch": 0.11,
"learning_rate": 9.733333333333333e-07,
"loss": 1.0468,
"step": 625
},
{
"epoch": 0.11,
"learning_rate": 9.677777777777777e-07,
"loss": 1.0424,
"step": 650
},
{
"epoch": 0.12,
"learning_rate": 9.62222222222222e-07,
"loss": 1.0406,
"step": 675
},
{
"epoch": 0.12,
"learning_rate": 9.566666666666667e-07,
"loss": 0.8936,
"step": 700
},
{
"epoch": 0.12,
"learning_rate": 9.511111111111111e-07,
"loss": 0.9655,
"step": 725
},
{
"epoch": 0.13,
"learning_rate": 9.455555555555556e-07,
"loss": 1.0059,
"step": 750
},
{
"epoch": 0.13,
"learning_rate": 9.399999999999999e-07,
"loss": 0.9234,
"step": 775
},
{
"epoch": 0.14,
"learning_rate": 9.344444444444444e-07,
"loss": 0.9642,
"step": 800
},
{
"epoch": 0.14,
"learning_rate": 9.288888888888888e-07,
"loss": 0.8986,
"step": 825
},
{
"epoch": 0.15,
"learning_rate": 9.233333333333333e-07,
"loss": 0.8999,
"step": 850
},
{
"epoch": 0.15,
"learning_rate": 9.177777777777777e-07,
"loss": 0.9481,
"step": 875
},
{
"epoch": 0.15,
"learning_rate": 9.122222222222222e-07,
"loss": 1.0157,
"step": 900
},
{
"epoch": 0.16,
"learning_rate": 9.066666666666665e-07,
"loss": 0.9704,
"step": 925
},
{
"epoch": 0.16,
"learning_rate": 9.01111111111111e-07,
"loss": 0.9003,
"step": 950
},
{
"epoch": 0.17,
"learning_rate": 8.955555555555555e-07,
"loss": 0.9299,
"step": 975
},
{
"epoch": 0.17,
"learning_rate": 8.9e-07,
"loss": 0.9274,
"step": 1000
},
{
"epoch": 0.17,
"eval_loss": 0.9543788433074951,
"eval_runtime": 2005.874,
"eval_samples_per_second": 1.222,
"eval_steps_per_second": 0.306,
"eval_wer": 67.9787303028338,
"step": 1000
},
{
"epoch": 0.18,
"learning_rate": 8.844444444444445e-07,
"loss": 0.9989,
"step": 1025
},
{
"epoch": 0.18,
"learning_rate": 8.788888888888889e-07,
"loss": 0.8707,
"step": 1050
},
{
"epoch": 0.18,
"learning_rate": 8.733333333333333e-07,
"loss": 1.0216,
"step": 1075
},
{
"epoch": 0.19,
"learning_rate": 8.677777777777777e-07,
"loss": 0.8865,
"step": 1100
},
{
"epoch": 0.19,
"learning_rate": 8.622222222222222e-07,
"loss": 0.9147,
"step": 1125
},
{
"epoch": 0.2,
"learning_rate": 8.566666666666667e-07,
"loss": 1.0797,
"step": 1150
},
{
"epoch": 0.2,
"learning_rate": 8.511111111111111e-07,
"loss": 1.0762,
"step": 1175
},
{
"epoch": 0.21,
"learning_rate": 8.455555555555555e-07,
"loss": 0.9013,
"step": 1200
},
{
"epoch": 0.21,
"learning_rate": 8.399999999999999e-07,
"loss": 0.9438,
"step": 1225
},
{
"epoch": 0.21,
"learning_rate": 8.344444444444444e-07,
"loss": 0.9928,
"step": 1250
},
{
"epoch": 0.22,
"learning_rate": 8.288888888888888e-07,
"loss": 0.8263,
"step": 1275
},
{
"epoch": 0.22,
"learning_rate": 8.233333333333333e-07,
"loss": 0.9346,
"step": 1300
},
{
"epoch": 0.23,
"learning_rate": 8.177777777777778e-07,
"loss": 0.8972,
"step": 1325
},
{
"epoch": 0.23,
"learning_rate": 8.122222222222221e-07,
"loss": 0.9257,
"step": 1350
},
{
"epoch": 0.24,
"learning_rate": 8.066666666666666e-07,
"loss": 0.9189,
"step": 1375
},
{
"epoch": 0.24,
"learning_rate": 8.01111111111111e-07,
"loss": 0.8484,
"step": 1400
},
{
"epoch": 0.24,
"learning_rate": 7.955555555555556e-07,
"loss": 0.9625,
"step": 1425
},
{
"epoch": 0.25,
"learning_rate": 7.9e-07,
"loss": 0.8788,
"step": 1450
},
{
"epoch": 0.25,
"learning_rate": 7.844444444444445e-07,
"loss": 0.9545,
"step": 1475
},
{
"epoch": 0.26,
"learning_rate": 7.788888888888889e-07,
"loss": 0.8733,
"step": 1500
},
{
"epoch": 0.26,
"learning_rate": 7.733333333333333e-07,
"loss": 1.0055,
"step": 1525
},
{
"epoch": 0.27,
"learning_rate": 7.677777777777778e-07,
"loss": 0.8434,
"step": 1550
},
{
"epoch": 0.27,
"learning_rate": 7.622222222222222e-07,
"loss": 0.9631,
"step": 1575
},
{
"epoch": 0.27,
"learning_rate": 7.566666666666667e-07,
"loss": 0.8942,
"step": 1600
},
{
"epoch": 0.28,
"learning_rate": 7.51111111111111e-07,
"loss": 0.823,
"step": 1625
},
{
"epoch": 0.28,
"learning_rate": 7.455555555555555e-07,
"loss": 0.9264,
"step": 1650
},
{
"epoch": 0.29,
"learning_rate": 7.4e-07,
"loss": 0.8994,
"step": 1675
},
{
"epoch": 0.29,
"learning_rate": 7.344444444444444e-07,
"loss": 0.833,
"step": 1700
},
{
"epoch": 0.3,
"learning_rate": 7.288888888888889e-07,
"loss": 0.8369,
"step": 1725
},
{
"epoch": 0.3,
"learning_rate": 7.233333333333333e-07,
"loss": 0.8402,
"step": 1750
},
{
"epoch": 0.3,
"learning_rate": 7.177777777777777e-07,
"loss": 0.9375,
"step": 1775
},
{
"epoch": 0.31,
"learning_rate": 7.122222222222221e-07,
"loss": 0.8242,
"step": 1800
},
{
"epoch": 0.31,
"learning_rate": 7.066666666666666e-07,
"loss": 0.952,
"step": 1825
},
{
"epoch": 0.32,
"learning_rate": 7.01111111111111e-07,
"loss": 0.8974,
"step": 1850
},
{
"epoch": 0.32,
"learning_rate": 6.955555555555556e-07,
"loss": 0.8254,
"step": 1875
},
{
"epoch": 0.33,
"learning_rate": 6.9e-07,
"loss": 0.8932,
"step": 1900
},
{
"epoch": 0.33,
"learning_rate": 6.844444444444444e-07,
"loss": 0.9254,
"step": 1925
},
{
"epoch": 0.33,
"learning_rate": 6.788888888888889e-07,
"loss": 0.9744,
"step": 1950
},
{
"epoch": 0.34,
"learning_rate": 6.733333333333333e-07,
"loss": 0.833,
"step": 1975
},
{
"epoch": 0.34,
"learning_rate": 6.677777777777778e-07,
"loss": 0.8826,
"step": 2000
},
{
"epoch": 0.34,
"eval_loss": 0.8867732286453247,
"eval_runtime": 1913.8237,
"eval_samples_per_second": 1.281,
"eval_steps_per_second": 0.32,
"eval_wer": 59.74320731470073,
"step": 2000
},
{
"epoch": 0.35,
"learning_rate": 6.622222222222222e-07,
"loss": 0.8571,
"step": 2025
},
{
"epoch": 0.35,
"learning_rate": 6.566666666666666e-07,
"loss": 0.9503,
"step": 2050
},
{
"epoch": 0.36,
"learning_rate": 6.511111111111111e-07,
"loss": 0.8916,
"step": 2075
},
{
"epoch": 0.36,
"learning_rate": 6.455555555555555e-07,
"loss": 0.8161,
"step": 2100
},
{
"epoch": 0.36,
"learning_rate": 6.4e-07,
"loss": 0.809,
"step": 2125
},
{
"epoch": 0.37,
"learning_rate": 6.344444444444444e-07,
"loss": 0.907,
"step": 2150
},
{
"epoch": 0.37,
"learning_rate": 6.288888888888889e-07,
"loss": 0.8803,
"step": 2175
},
{
"epoch": 0.38,
"learning_rate": 6.233333333333332e-07,
"loss": 0.8582,
"step": 2200
},
{
"epoch": 0.38,
"learning_rate": 6.177777777777777e-07,
"loss": 0.8639,
"step": 2225
},
{
"epoch": 0.39,
"learning_rate": 6.122222222222222e-07,
"loss": 0.7709,
"step": 2250
},
{
"epoch": 0.39,
"learning_rate": 6.066666666666666e-07,
"loss": 0.7636,
"step": 2275
},
{
"epoch": 0.4,
"learning_rate": 6.011111111111112e-07,
"loss": 0.9221,
"step": 2300
},
{
"epoch": 0.4,
"learning_rate": 5.955555555555555e-07,
"loss": 0.9319,
"step": 2325
},
{
"epoch": 0.4,
"learning_rate": 5.9e-07,
"loss": 0.9681,
"step": 2350
},
{
"epoch": 0.41,
"learning_rate": 5.844444444444444e-07,
"loss": 0.9214,
"step": 2375
},
{
"epoch": 0.41,
"learning_rate": 5.788888888888889e-07,
"loss": 0.8284,
"step": 2400
},
{
"epoch": 0.42,
"learning_rate": 5.733333333333334e-07,
"loss": 0.863,
"step": 2425
},
{
"epoch": 0.42,
"learning_rate": 5.677777777777778e-07,
"loss": 0.8422,
"step": 2450
},
{
"epoch": 0.43,
"learning_rate": 5.622222222222222e-07,
"loss": 0.868,
"step": 2475
},
{
"epoch": 0.43,
"learning_rate": 5.566666666666666e-07,
"loss": 0.8797,
"step": 2500
},
{
"epoch": 0.43,
"learning_rate": 5.511111111111111e-07,
"loss": 0.8566,
"step": 2525
},
{
"epoch": 0.44,
"learning_rate": 5.455555555555555e-07,
"loss": 0.8576,
"step": 2550
},
{
"epoch": 0.44,
"learning_rate": 5.4e-07,
"loss": 0.8294,
"step": 2575
},
{
"epoch": 0.45,
"learning_rate": 5.344444444444445e-07,
"loss": 0.808,
"step": 2600
},
{
"epoch": 0.45,
"learning_rate": 5.288888888888888e-07,
"loss": 0.8659,
"step": 2625
},
{
"epoch": 0.46,
"learning_rate": 5.233333333333333e-07,
"loss": 0.9605,
"step": 2650
},
{
"epoch": 0.46,
"learning_rate": 5.177777777777777e-07,
"loss": 0.8541,
"step": 2675
},
{
"epoch": 0.46,
"learning_rate": 5.122222222222222e-07,
"loss": 0.8927,
"step": 2700
},
{
"epoch": 0.47,
"learning_rate": 5.066666666666667e-07,
"loss": 0.9116,
"step": 2725
},
{
"epoch": 0.47,
"learning_rate": 5.011111111111111e-07,
"loss": 0.8917,
"step": 2750
},
{
"epoch": 0.48,
"learning_rate": 4.955555555555556e-07,
"loss": 0.9124,
"step": 2775
},
{
"epoch": 0.48,
"learning_rate": 4.9e-07,
"loss": 0.7529,
"step": 2800
},
{
"epoch": 0.49,
"learning_rate": 4.844444444444445e-07,
"loss": 0.8252,
"step": 2825
},
{
"epoch": 0.49,
"learning_rate": 4.788888888888889e-07,
"loss": 0.8791,
"step": 2850
},
{
"epoch": 0.49,
"learning_rate": 4.733333333333333e-07,
"loss": 0.9518,
"step": 2875
},
{
"epoch": 0.5,
"learning_rate": 4.677777777777778e-07,
"loss": 0.7611,
"step": 2900
},
{
"epoch": 0.5,
"learning_rate": 4.622222222222222e-07,
"loss": 0.9263,
"step": 2925
},
{
"epoch": 0.51,
"learning_rate": 4.5666666666666665e-07,
"loss": 0.9538,
"step": 2950
},
{
"epoch": 0.51,
"learning_rate": 4.511111111111111e-07,
"loss": 0.8454,
"step": 2975
},
{
"epoch": 0.52,
"learning_rate": 4.455555555555555e-07,
"loss": 0.8188,
"step": 3000
},
{
"epoch": 0.52,
"eval_loss": 0.8566001057624817,
"eval_runtime": 1864.0762,
"eval_samples_per_second": 1.315,
"eval_steps_per_second": 0.329,
"eval_wer": 57.525452305297975,
"step": 3000
},
{
"epoch": 0.52,
"learning_rate": 4.3999999999999997e-07,
"loss": 0.8941,
"step": 3025
},
{
"epoch": 0.52,
"learning_rate": 4.344444444444444e-07,
"loss": 0.815,
"step": 3050
},
{
"epoch": 0.53,
"learning_rate": 4.2888888888888883e-07,
"loss": 0.7275,
"step": 3075
},
{
"epoch": 0.53,
"learning_rate": 4.2333333333333334e-07,
"loss": 0.8606,
"step": 3100
},
{
"epoch": 0.54,
"learning_rate": 4.177777777777778e-07,
"loss": 0.8937,
"step": 3125
},
{
"epoch": 0.54,
"learning_rate": 4.122222222222222e-07,
"loss": 0.8924,
"step": 3150
},
{
"epoch": 0.55,
"learning_rate": 4.0666666666666666e-07,
"loss": 0.8555,
"step": 3175
},
{
"epoch": 0.55,
"learning_rate": 4.0111111111111106e-07,
"loss": 0.9383,
"step": 3200
},
{
"epoch": 0.55,
"learning_rate": 3.955555555555555e-07,
"loss": 0.8445,
"step": 3225
},
{
"epoch": 0.56,
"learning_rate": 3.8999999999999997e-07,
"loss": 0.7983,
"step": 3250
},
{
"epoch": 0.56,
"learning_rate": 3.8444444444444443e-07,
"loss": 0.873,
"step": 3275
},
{
"epoch": 0.57,
"learning_rate": 3.788888888888889e-07,
"loss": 0.767,
"step": 3300
},
{
"epoch": 0.57,
"learning_rate": 3.7333333333333334e-07,
"loss": 0.8215,
"step": 3325
},
{
"epoch": 0.58,
"learning_rate": 3.6777777777777774e-07,
"loss": 0.7633,
"step": 3350
},
{
"epoch": 0.58,
"learning_rate": 3.622222222222222e-07,
"loss": 0.8075,
"step": 3375
},
{
"epoch": 0.58,
"learning_rate": 3.5666666666666666e-07,
"loss": 0.7566,
"step": 3400
},
{
"epoch": 0.59,
"learning_rate": 3.5111111111111106e-07,
"loss": 0.92,
"step": 3425
},
{
"epoch": 0.59,
"learning_rate": 3.4555555555555557e-07,
"loss": 0.87,
"step": 3450
},
{
"epoch": 0.6,
"learning_rate": 3.4000000000000003e-07,
"loss": 0.9613,
"step": 3475
},
{
"epoch": 0.6,
"learning_rate": 3.3444444444444443e-07,
"loss": 0.8107,
"step": 3500
},
{
"epoch": 0.61,
"learning_rate": 3.288888888888889e-07,
"loss": 0.8787,
"step": 3525
},
{
"epoch": 0.61,
"learning_rate": 3.233333333333333e-07,
"loss": 0.7038,
"step": 3550
},
{
"epoch": 0.61,
"learning_rate": 3.1777777777777775e-07,
"loss": 0.7843,
"step": 3575
},
{
"epoch": 0.62,
"learning_rate": 3.122222222222222e-07,
"loss": 0.8556,
"step": 3600
},
{
"epoch": 0.62,
"learning_rate": 3.066666666666666e-07,
"loss": 0.8448,
"step": 3625
},
{
"epoch": 0.63,
"learning_rate": 3.011111111111111e-07,
"loss": 0.8249,
"step": 3650
},
{
"epoch": 0.63,
"learning_rate": 2.9555555555555557e-07,
"loss": 0.8616,
"step": 3675
},
{
"epoch": 0.64,
"learning_rate": 2.9e-07,
"loss": 0.9052,
"step": 3700
},
{
"epoch": 0.64,
"learning_rate": 2.8444444444444443e-07,
"loss": 0.7239,
"step": 3725
},
{
"epoch": 0.64,
"learning_rate": 2.788888888888889e-07,
"loss": 0.9165,
"step": 3750
},
{
"epoch": 0.65,
"learning_rate": 2.733333333333333e-07,
"loss": 0.8461,
"step": 3775
},
{
"epoch": 0.65,
"learning_rate": 2.6777777777777775e-07,
"loss": 0.8162,
"step": 3800
},
{
"epoch": 0.66,
"learning_rate": 2.6222222222222226e-07,
"loss": 0.8413,
"step": 3825
},
{
"epoch": 0.66,
"learning_rate": 2.5666666666666666e-07,
"loss": 0.8478,
"step": 3850
},
{
"epoch": 0.67,
"learning_rate": 2.511111111111111e-07,
"loss": 0.9064,
"step": 3875
},
{
"epoch": 0.67,
"learning_rate": 2.455555555555555e-07,
"loss": 0.8219,
"step": 3900
},
{
"epoch": 0.67,
"learning_rate": 2.4e-07,
"loss": 0.7095,
"step": 3925
},
{
"epoch": 0.68,
"learning_rate": 2.3444444444444444e-07,
"loss": 0.9094,
"step": 3950
},
{
"epoch": 0.68,
"learning_rate": 2.288888888888889e-07,
"loss": 0.7578,
"step": 3975
},
{
"epoch": 0.69,
"learning_rate": 2.2333333333333332e-07,
"loss": 0.8253,
"step": 4000
},
{
"epoch": 0.69,
"eval_loss": 0.8424507975578308,
"eval_runtime": 1839.3907,
"eval_samples_per_second": 1.333,
"eval_steps_per_second": 0.333,
"eval_wer": 53.80973996498282,
"step": 4000
},
{
"epoch": 0.69,
"learning_rate": 2.1777777777777775e-07,
"loss": 0.8382,
"step": 4025
},
{
"epoch": 0.7,
"learning_rate": 2.122222222222222e-07,
"loss": 0.8846,
"step": 4050
},
{
"epoch": 0.7,
"learning_rate": 2.0666666666666666e-07,
"loss": 0.8043,
"step": 4075
},
{
"epoch": 0.7,
"learning_rate": 2.011111111111111e-07,
"loss": 0.7558,
"step": 4100
},
{
"epoch": 0.71,
"learning_rate": 1.9555555555555555e-07,
"loss": 0.8159,
"step": 4125
},
{
"epoch": 0.71,
"learning_rate": 1.8999999999999998e-07,
"loss": 0.837,
"step": 4150
},
{
"epoch": 0.72,
"learning_rate": 1.8444444444444444e-07,
"loss": 0.8796,
"step": 4175
},
{
"epoch": 0.72,
"learning_rate": 1.7888888888888887e-07,
"loss": 0.7855,
"step": 4200
},
{
"epoch": 0.73,
"learning_rate": 1.7333333333333332e-07,
"loss": 0.8801,
"step": 4225
},
{
"epoch": 0.73,
"learning_rate": 1.6777777777777778e-07,
"loss": 0.8426,
"step": 4250
},
{
"epoch": 0.73,
"learning_rate": 1.622222222222222e-07,
"loss": 0.9545,
"step": 4275
},
{
"epoch": 0.74,
"learning_rate": 1.5666666666666667e-07,
"loss": 0.7671,
"step": 4300
},
{
"epoch": 0.74,
"learning_rate": 1.511111111111111e-07,
"loss": 0.74,
"step": 4325
},
{
"epoch": 0.75,
"learning_rate": 1.4555555555555555e-07,
"loss": 0.9471,
"step": 4350
},
{
"epoch": 0.75,
"learning_rate": 1.4e-07,
"loss": 0.7438,
"step": 4375
},
{
"epoch": 0.76,
"learning_rate": 1.3444444444444444e-07,
"loss": 0.9543,
"step": 4400
},
{
"epoch": 0.76,
"learning_rate": 1.2888888888888887e-07,
"loss": 0.9077,
"step": 4425
},
{
"epoch": 0.76,
"learning_rate": 1.2333333333333333e-07,
"loss": 0.8108,
"step": 4450
},
{
"epoch": 0.77,
"learning_rate": 1.1777777777777777e-07,
"loss": 0.9617,
"step": 4475
},
{
"epoch": 0.77,
"learning_rate": 1.1222222222222221e-07,
"loss": 0.8158,
"step": 4500
},
{
"epoch": 0.78,
"learning_rate": 1.0666666666666667e-07,
"loss": 0.7763,
"step": 4525
},
{
"epoch": 0.78,
"learning_rate": 1.011111111111111e-07,
"loss": 0.8154,
"step": 4550
},
{
"epoch": 0.79,
"learning_rate": 9.555555555555556e-08,
"loss": 0.8584,
"step": 4575
},
{
"epoch": 0.79,
"learning_rate": 9e-08,
"loss": 0.8362,
"step": 4600
},
{
"epoch": 0.79,
"learning_rate": 8.444444444444444e-08,
"loss": 0.8365,
"step": 4625
},
{
"epoch": 0.8,
"learning_rate": 7.888888888888889e-08,
"loss": 0.8727,
"step": 4650
},
{
"epoch": 0.8,
"learning_rate": 7.333333333333333e-08,
"loss": 0.8842,
"step": 4675
},
{
"epoch": 0.81,
"learning_rate": 6.777777777777778e-08,
"loss": 0.9152,
"step": 4700
},
{
"epoch": 0.81,
"learning_rate": 6.222222222222221e-08,
"loss": 0.8215,
"step": 4725
},
{
"epoch": 0.82,
"learning_rate": 5.666666666666666e-08,
"loss": 0.7567,
"step": 4750
},
{
"epoch": 0.82,
"learning_rate": 5.1111111111111114e-08,
"loss": 0.934,
"step": 4775
},
{
"epoch": 0.82,
"learning_rate": 4.555555555555556e-08,
"loss": 0.7867,
"step": 4800
},
{
"epoch": 0.83,
"learning_rate": 4e-08,
"loss": 0.7707,
"step": 4825
},
{
"epoch": 0.83,
"learning_rate": 3.4444444444444444e-08,
"loss": 0.8075,
"step": 4850
},
{
"epoch": 0.84,
"learning_rate": 2.8888888888888887e-08,
"loss": 0.794,
"step": 4875
},
{
"epoch": 0.84,
"learning_rate": 2.3333333333333334e-08,
"loss": 0.8668,
"step": 4900
},
{
"epoch": 0.85,
"learning_rate": 1.7777777777777777e-08,
"loss": 0.8028,
"step": 4925
},
{
"epoch": 0.85,
"learning_rate": 1.2222222222222222e-08,
"loss": 0.7995,
"step": 4950
},
{
"epoch": 0.85,
"learning_rate": 6.666666666666667e-09,
"loss": 0.7102,
"step": 4975
},
{
"epoch": 0.86,
"learning_rate": 1.111111111111111e-09,
"loss": 0.8112,
"step": 5000
},
{
"epoch": 0.86,
"eval_loss": 0.8372716307640076,
"eval_runtime": 1890.621,
"eval_samples_per_second": 1.296,
"eval_steps_per_second": 0.324,
"eval_wer": 56.44251345567732,
"step": 5000
},
{
"epoch": 0.86,
"step": 5000,
"total_flos": 1.15434160128e+19,
"train_loss": 0.955653938293457,
"train_runtime": 35094.2694,
"train_samples_per_second": 1.14,
"train_steps_per_second": 0.142
}
],
"max_steps": 5000,
"num_train_epochs": 1,
"total_flos": 1.15434160128e+19,
"trial_name": null,
"trial_params": null
}