whisper-md-hu / trainer_state.json
emilios's picture
End of training
4de691e
{
"best_metric": 18.642158316039133,
"best_model_checkpoint": "./checkpoint-8000",
"epoch": 10.493179433368311,
"global_step": 10000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.03,
"learning_rate": 1.4921499816584212e-06,
"loss": 0.8781,
"step": 25
},
{
"epoch": 0.05,
"learning_rate": 1.8585955256072678e-06,
"loss": 0.4617,
"step": 50
},
{
"epoch": 0.08,
"learning_rate": 2.0644903353287132e-06,
"loss": 0.3577,
"step": 75
},
{
"epoch": 0.1,
"learning_rate": 2.208366596599153e-06,
"loss": 0.3052,
"step": 100
},
{
"epoch": 0.13,
"learning_rate": 2.319062265222936e-06,
"loss": 0.2758,
"step": 125
},
{
"epoch": 0.16,
"learning_rate": 2.409049375798488e-06,
"loss": 0.2643,
"step": 150
},
{
"epoch": 0.18,
"learning_rate": 2.4848684238610323e-06,
"loss": 0.2346,
"step": 175
},
{
"epoch": 0.21,
"learning_rate": 2.5503798365399707e-06,
"loss": 0.2365,
"step": 200
},
{
"epoch": 0.24,
"learning_rate": 2.608053780532331e-06,
"loss": 0.2304,
"step": 225
},
{
"epoch": 0.26,
"learning_rate": 2.65956674147807e-06,
"loss": 0.2095,
"step": 250
},
{
"epoch": 0.29,
"learning_rate": 2.7061088861190967e-06,
"loss": 0.1963,
"step": 275
},
{
"epoch": 0.31,
"learning_rate": 2.748555684588509e-06,
"loss": 0.1967,
"step": 300
},
{
"epoch": 0.34,
"learning_rate": 2.787569925934908e-06,
"loss": 0.179,
"step": 325
},
{
"epoch": 0.37,
"learning_rate": 2.8236654768779034e-06,
"loss": 0.1743,
"step": 350
},
{
"epoch": 0.39,
"learning_rate": 2.8572488050095434e-06,
"loss": 0.1694,
"step": 375
},
{
"epoch": 0.42,
"learning_rate": 2.8886469682005057e-06,
"loss": 0.1765,
"step": 400
},
{
"epoch": 0.45,
"learning_rate": 2.91812704114235e-06,
"loss": 0.1707,
"step": 425
},
{
"epoch": 0.47,
"learning_rate": 2.94590994217138e-06,
"loss": 0.1546,
"step": 450
},
{
"epoch": 0.5,
"learning_rate": 2.972180492193256e-06,
"loss": 0.1443,
"step": 475
},
{
"epoch": 0.52,
"learning_rate": 2.997094874416736e-06,
"loss": 0.1555,
"step": 500
},
{
"epoch": 0.55,
"learning_rate": 3e-06,
"loss": 0.1434,
"step": 525
},
{
"epoch": 0.58,
"learning_rate": 3e-06,
"loss": 0.1453,
"step": 550
},
{
"epoch": 0.6,
"learning_rate": 3e-06,
"loss": 0.1255,
"step": 575
},
{
"epoch": 0.63,
"learning_rate": 3e-06,
"loss": 0.1292,
"step": 600
},
{
"epoch": 0.66,
"learning_rate": 3e-06,
"loss": 0.1306,
"step": 625
},
{
"epoch": 0.68,
"learning_rate": 3e-06,
"loss": 0.1223,
"step": 650
},
{
"epoch": 0.71,
"learning_rate": 3e-06,
"loss": 0.1358,
"step": 675
},
{
"epoch": 0.73,
"learning_rate": 3e-06,
"loss": 0.1279,
"step": 700
},
{
"epoch": 0.76,
"learning_rate": 3e-06,
"loss": 0.1192,
"step": 725
},
{
"epoch": 0.79,
"learning_rate": 3e-06,
"loss": 0.1135,
"step": 750
},
{
"epoch": 0.81,
"learning_rate": 3e-06,
"loss": 0.1076,
"step": 775
},
{
"epoch": 0.84,
"learning_rate": 3e-06,
"loss": 0.1105,
"step": 800
},
{
"epoch": 0.87,
"learning_rate": 3e-06,
"loss": 0.1152,
"step": 825
},
{
"epoch": 0.89,
"learning_rate": 3e-06,
"loss": 0.1025,
"step": 850
},
{
"epoch": 0.92,
"learning_rate": 3e-06,
"loss": 0.1076,
"step": 875
},
{
"epoch": 0.94,
"learning_rate": 3e-06,
"loss": 0.0985,
"step": 900
},
{
"epoch": 0.97,
"learning_rate": 3e-06,
"loss": 0.0849,
"step": 925
},
{
"epoch": 1.0,
"learning_rate": 3e-06,
"loss": 0.0962,
"step": 950
},
{
"epoch": 1.02,
"learning_rate": 3e-06,
"loss": 0.0615,
"step": 975
},
{
"epoch": 1.05,
"learning_rate": 3e-06,
"loss": 0.0621,
"step": 1000
},
{
"epoch": 1.05,
"eval_loss": 0.26904296875,
"eval_runtime": 820.8166,
"eval_samples_per_second": 5.642,
"eval_steps_per_second": 0.353,
"eval_wer": 20.509931811443817,
"step": 1000
},
{
"epoch": 1.08,
"learning_rate": 3e-06,
"loss": 0.0612,
"step": 1025
},
{
"epoch": 1.1,
"learning_rate": 3e-06,
"loss": 0.0609,
"step": 1050
},
{
"epoch": 1.13,
"learning_rate": 3e-06,
"loss": 0.0501,
"step": 1075
},
{
"epoch": 1.15,
"learning_rate": 3e-06,
"loss": 0.0586,
"step": 1100
},
{
"epoch": 1.18,
"learning_rate": 3e-06,
"loss": 0.0573,
"step": 1125
},
{
"epoch": 1.21,
"learning_rate": 3e-06,
"loss": 0.0577,
"step": 1150
},
{
"epoch": 1.23,
"learning_rate": 3e-06,
"loss": 0.052,
"step": 1175
},
{
"epoch": 1.26,
"learning_rate": 3e-06,
"loss": 0.0557,
"step": 1200
},
{
"epoch": 1.29,
"learning_rate": 3e-06,
"loss": 0.0553,
"step": 1225
},
{
"epoch": 1.31,
"learning_rate": 3e-06,
"loss": 0.0536,
"step": 1250
},
{
"epoch": 1.34,
"learning_rate": 3e-06,
"loss": 0.0493,
"step": 1275
},
{
"epoch": 1.36,
"learning_rate": 3e-06,
"loss": 0.0509,
"step": 1300
},
{
"epoch": 1.39,
"learning_rate": 3e-06,
"loss": 0.0527,
"step": 1325
},
{
"epoch": 1.42,
"learning_rate": 3e-06,
"loss": 0.049,
"step": 1350
},
{
"epoch": 1.44,
"learning_rate": 3e-06,
"loss": 0.0489,
"step": 1375
},
{
"epoch": 1.47,
"learning_rate": 3e-06,
"loss": 0.0514,
"step": 1400
},
{
"epoch": 1.5,
"learning_rate": 3e-06,
"loss": 0.0413,
"step": 1425
},
{
"epoch": 1.52,
"learning_rate": 3e-06,
"loss": 0.0472,
"step": 1450
},
{
"epoch": 1.55,
"learning_rate": 3e-06,
"loss": 0.0446,
"step": 1475
},
{
"epoch": 1.57,
"learning_rate": 3e-06,
"loss": 0.0438,
"step": 1500
},
{
"epoch": 1.6,
"learning_rate": 3e-06,
"loss": 0.0496,
"step": 1525
},
{
"epoch": 1.63,
"learning_rate": 3e-06,
"loss": 0.049,
"step": 1550
},
{
"epoch": 1.65,
"learning_rate": 3e-06,
"loss": 0.0434,
"step": 1575
},
{
"epoch": 1.68,
"learning_rate": 3e-06,
"loss": 0.0434,
"step": 1600
},
{
"epoch": 1.71,
"learning_rate": 3e-06,
"loss": 0.0412,
"step": 1625
},
{
"epoch": 1.73,
"learning_rate": 3e-06,
"loss": 0.0444,
"step": 1650
},
{
"epoch": 1.76,
"learning_rate": 3e-06,
"loss": 0.0442,
"step": 1675
},
{
"epoch": 1.78,
"learning_rate": 3e-06,
"loss": 0.0444,
"step": 1700
},
{
"epoch": 1.81,
"learning_rate": 3e-06,
"loss": 0.0418,
"step": 1725
},
{
"epoch": 1.84,
"learning_rate": 3e-06,
"loss": 0.0409,
"step": 1750
},
{
"epoch": 1.86,
"learning_rate": 3e-06,
"loss": 0.0423,
"step": 1775
},
{
"epoch": 1.89,
"learning_rate": 3e-06,
"loss": 0.0425,
"step": 1800
},
{
"epoch": 1.92,
"learning_rate": 3e-06,
"loss": 0.0383,
"step": 1825
},
{
"epoch": 1.94,
"learning_rate": 3e-06,
"loss": 0.0417,
"step": 1850
},
{
"epoch": 1.97,
"learning_rate": 3e-06,
"loss": 0.0357,
"step": 1875
},
{
"epoch": 1.99,
"learning_rate": 3e-06,
"loss": 0.0391,
"step": 1900
},
{
"epoch": 2.02,
"learning_rate": 3e-06,
"loss": 0.021,
"step": 1925
},
{
"epoch": 2.05,
"learning_rate": 3e-06,
"loss": 0.0178,
"step": 1950
},
{
"epoch": 2.07,
"learning_rate": 3e-06,
"loss": 0.0202,
"step": 1975
},
{
"epoch": 2.1,
"learning_rate": 3e-06,
"loss": 0.0174,
"step": 2000
},
{
"epoch": 2.1,
"eval_loss": 0.2705078125,
"eval_runtime": 820.5458,
"eval_samples_per_second": 5.644,
"eval_steps_per_second": 0.353,
"eval_wer": 19.22917284316632,
"step": 2000
},
{
"epoch": 2.12,
"learning_rate": 3e-06,
"loss": 0.0171,
"step": 2025
},
{
"epoch": 2.15,
"learning_rate": 3e-06,
"loss": 0.0167,
"step": 2050
},
{
"epoch": 2.18,
"learning_rate": 3e-06,
"loss": 0.0176,
"step": 2075
},
{
"epoch": 2.2,
"learning_rate": 3e-06,
"loss": 0.0165,
"step": 2100
},
{
"epoch": 2.23,
"learning_rate": 3e-06,
"loss": 0.0203,
"step": 2125
},
{
"epoch": 2.26,
"learning_rate": 3e-06,
"loss": 0.0167,
"step": 2150
},
{
"epoch": 2.28,
"learning_rate": 3e-06,
"loss": 0.0156,
"step": 2175
},
{
"epoch": 2.31,
"learning_rate": 3e-06,
"loss": 0.0168,
"step": 2200
},
{
"epoch": 2.33,
"learning_rate": 3e-06,
"loss": 0.0162,
"step": 2225
},
{
"epoch": 2.36,
"learning_rate": 3e-06,
"loss": 0.0168,
"step": 2250
},
{
"epoch": 2.39,
"learning_rate": 3e-06,
"loss": 0.0181,
"step": 2275
},
{
"epoch": 2.41,
"learning_rate": 3e-06,
"loss": 0.0175,
"step": 2300
},
{
"epoch": 2.44,
"learning_rate": 3e-06,
"loss": 0.0155,
"step": 2325
},
{
"epoch": 2.47,
"learning_rate": 3e-06,
"loss": 0.0168,
"step": 2350
},
{
"epoch": 2.49,
"learning_rate": 3e-06,
"loss": 0.016,
"step": 2375
},
{
"epoch": 2.52,
"learning_rate": 3e-06,
"loss": 0.0153,
"step": 2400
},
{
"epoch": 2.54,
"learning_rate": 3e-06,
"loss": 0.0159,
"step": 2425
},
{
"epoch": 2.57,
"learning_rate": 3e-06,
"loss": 0.015,
"step": 2450
},
{
"epoch": 2.6,
"learning_rate": 3e-06,
"loss": 0.0154,
"step": 2475
},
{
"epoch": 2.62,
"learning_rate": 3e-06,
"loss": 0.0154,
"step": 2500
},
{
"epoch": 2.65,
"learning_rate": 3e-06,
"loss": 0.0176,
"step": 2525
},
{
"epoch": 2.68,
"learning_rate": 3e-06,
"loss": 0.0161,
"step": 2550
},
{
"epoch": 2.7,
"learning_rate": 3e-06,
"loss": 0.0123,
"step": 2575
},
{
"epoch": 2.73,
"learning_rate": 3e-06,
"loss": 0.0163,
"step": 2600
},
{
"epoch": 2.75,
"learning_rate": 3e-06,
"loss": 0.0173,
"step": 2625
},
{
"epoch": 2.78,
"learning_rate": 3e-06,
"loss": 0.0145,
"step": 2650
},
{
"epoch": 2.81,
"learning_rate": 3e-06,
"loss": 0.014,
"step": 2675
},
{
"epoch": 2.83,
"learning_rate": 3e-06,
"loss": 0.0145,
"step": 2700
},
{
"epoch": 2.86,
"learning_rate": 3e-06,
"loss": 0.0179,
"step": 2725
},
{
"epoch": 2.89,
"learning_rate": 3e-06,
"loss": 0.0168,
"step": 2750
},
{
"epoch": 2.91,
"learning_rate": 3e-06,
"loss": 0.0123,
"step": 2775
},
{
"epoch": 2.94,
"learning_rate": 3e-06,
"loss": 0.0144,
"step": 2800
},
{
"epoch": 2.96,
"learning_rate": 3e-06,
"loss": 0.015,
"step": 2825
},
{
"epoch": 2.99,
"learning_rate": 3e-06,
"loss": 0.015,
"step": 2850
},
{
"epoch": 3.02,
"learning_rate": 3e-06,
"loss": 0.0101,
"step": 2875
},
{
"epoch": 3.04,
"learning_rate": 3e-06,
"loss": 0.0075,
"step": 2900
},
{
"epoch": 3.07,
"learning_rate": 3e-06,
"loss": 0.0069,
"step": 2925
},
{
"epoch": 3.1,
"learning_rate": 3e-06,
"loss": 0.0059,
"step": 2950
},
{
"epoch": 3.12,
"learning_rate": 3e-06,
"loss": 0.0062,
"step": 2975
},
{
"epoch": 3.15,
"learning_rate": 3e-06,
"loss": 0.006,
"step": 3000
},
{
"epoch": 3.15,
"eval_loss": 0.29541015625,
"eval_runtime": 821.5028,
"eval_samples_per_second": 5.637,
"eval_steps_per_second": 0.353,
"eval_wer": 18.98903053661429,
"step": 3000
},
{
"epoch": 3.17,
"learning_rate": 3e-06,
"loss": 0.006,
"step": 3025
},
{
"epoch": 3.2,
"learning_rate": 3e-06,
"loss": 0.0064,
"step": 3050
},
{
"epoch": 3.23,
"learning_rate": 3e-06,
"loss": 0.0056,
"step": 3075
},
{
"epoch": 3.25,
"learning_rate": 3e-06,
"loss": 0.0056,
"step": 3100
},
{
"epoch": 3.28,
"learning_rate": 3e-06,
"loss": 0.0064,
"step": 3125
},
{
"epoch": 3.31,
"learning_rate": 3e-06,
"loss": 0.0079,
"step": 3150
},
{
"epoch": 3.33,
"learning_rate": 3e-06,
"loss": 0.0064,
"step": 3175
},
{
"epoch": 3.36,
"learning_rate": 3e-06,
"loss": 0.0065,
"step": 3200
},
{
"epoch": 3.38,
"learning_rate": 3e-06,
"loss": 0.0053,
"step": 3225
},
{
"epoch": 3.41,
"learning_rate": 3e-06,
"loss": 0.007,
"step": 3250
},
{
"epoch": 3.44,
"learning_rate": 3e-06,
"loss": 0.0061,
"step": 3275
},
{
"epoch": 3.46,
"learning_rate": 3e-06,
"loss": 0.0067,
"step": 3300
},
{
"epoch": 3.49,
"learning_rate": 3e-06,
"loss": 0.0074,
"step": 3325
},
{
"epoch": 3.52,
"learning_rate": 3e-06,
"loss": 0.0063,
"step": 3350
},
{
"epoch": 3.54,
"learning_rate": 3e-06,
"loss": 0.0062,
"step": 3375
},
{
"epoch": 3.57,
"learning_rate": 3e-06,
"loss": 0.0072,
"step": 3400
},
{
"epoch": 3.59,
"learning_rate": 3e-06,
"loss": 0.0061,
"step": 3425
},
{
"epoch": 3.62,
"learning_rate": 3e-06,
"loss": 0.0059,
"step": 3450
},
{
"epoch": 3.65,
"learning_rate": 3e-06,
"loss": 0.006,
"step": 3475
},
{
"epoch": 3.67,
"learning_rate": 3e-06,
"loss": 0.0062,
"step": 3500
},
{
"epoch": 3.7,
"learning_rate": 3e-06,
"loss": 0.0062,
"step": 3525
},
{
"epoch": 3.73,
"learning_rate": 3e-06,
"loss": 0.0056,
"step": 3550
},
{
"epoch": 3.75,
"learning_rate": 3e-06,
"loss": 0.0062,
"step": 3575
},
{
"epoch": 3.78,
"learning_rate": 3e-06,
"loss": 0.0062,
"step": 3600
},
{
"epoch": 3.8,
"learning_rate": 3e-06,
"loss": 0.0066,
"step": 3625
},
{
"epoch": 3.83,
"learning_rate": 3e-06,
"loss": 0.0067,
"step": 3650
},
{
"epoch": 3.86,
"learning_rate": 3e-06,
"loss": 0.0065,
"step": 3675
},
{
"epoch": 3.88,
"learning_rate": 3e-06,
"loss": 0.0069,
"step": 3700
},
{
"epoch": 3.91,
"learning_rate": 3e-06,
"loss": 0.0068,
"step": 3725
},
{
"epoch": 3.93,
"learning_rate": 3e-06,
"loss": 0.0063,
"step": 3750
},
{
"epoch": 3.96,
"learning_rate": 3e-06,
"loss": 0.0062,
"step": 3775
},
{
"epoch": 3.99,
"learning_rate": 3e-06,
"loss": 0.0065,
"step": 3800
},
{
"epoch": 4.01,
"learning_rate": 3e-06,
"loss": 0.005,
"step": 3825
},
{
"epoch": 4.04,
"learning_rate": 3e-06,
"loss": 0.0034,
"step": 3850
},
{
"epoch": 4.07,
"learning_rate": 3e-06,
"loss": 0.0032,
"step": 3875
},
{
"epoch": 4.09,
"learning_rate": 3e-06,
"loss": 0.0037,
"step": 3900
},
{
"epoch": 4.12,
"learning_rate": 3e-06,
"loss": 0.0028,
"step": 3925
},
{
"epoch": 4.14,
"learning_rate": 3e-06,
"loss": 0.0031,
"step": 3950
},
{
"epoch": 4.17,
"learning_rate": 3e-06,
"loss": 0.0031,
"step": 3975
},
{
"epoch": 4.2,
"learning_rate": 3e-06,
"loss": 0.0028,
"step": 4000
},
{
"epoch": 4.2,
"eval_loss": 0.309326171875,
"eval_runtime": 821.563,
"eval_samples_per_second": 5.637,
"eval_steps_per_second": 0.353,
"eval_wer": 18.802253187073823,
"step": 4000
},
{
"epoch": 4.22,
"learning_rate": 3e-06,
"loss": 0.0029,
"step": 4025
},
{
"epoch": 4.25,
"learning_rate": 3e-06,
"loss": 0.0035,
"step": 4050
},
{
"epoch": 4.28,
"learning_rate": 3e-06,
"loss": 0.0028,
"step": 4075
},
{
"epoch": 4.3,
"learning_rate": 3e-06,
"loss": 0.003,
"step": 4100
},
{
"epoch": 4.33,
"learning_rate": 3e-06,
"loss": 0.0035,
"step": 4125
},
{
"epoch": 4.35,
"learning_rate": 3e-06,
"loss": 0.0029,
"step": 4150
},
{
"epoch": 4.38,
"learning_rate": 3e-06,
"loss": 0.0036,
"step": 4175
},
{
"epoch": 4.41,
"learning_rate": 3e-06,
"loss": 0.003,
"step": 4200
},
{
"epoch": 4.43,
"learning_rate": 3e-06,
"loss": 0.0032,
"step": 4225
},
{
"epoch": 4.46,
"learning_rate": 3e-06,
"loss": 0.0038,
"step": 4250
},
{
"epoch": 4.49,
"learning_rate": 3e-06,
"loss": 0.0029,
"step": 4275
},
{
"epoch": 4.51,
"learning_rate": 3e-06,
"loss": 0.0047,
"step": 4300
},
{
"epoch": 4.54,
"learning_rate": 3e-06,
"loss": 0.0031,
"step": 4325
},
{
"epoch": 4.56,
"learning_rate": 3e-06,
"loss": 0.0036,
"step": 4350
},
{
"epoch": 4.59,
"learning_rate": 3e-06,
"loss": 0.0042,
"step": 4375
},
{
"epoch": 4.62,
"learning_rate": 3e-06,
"loss": 0.004,
"step": 4400
},
{
"epoch": 4.64,
"learning_rate": 3e-06,
"loss": 0.0033,
"step": 4425
},
{
"epoch": 4.67,
"learning_rate": 3e-06,
"loss": 0.0047,
"step": 4450
},
{
"epoch": 4.7,
"learning_rate": 3e-06,
"loss": 0.0031,
"step": 4475
},
{
"epoch": 4.72,
"learning_rate": 3e-06,
"loss": 0.0032,
"step": 4500
},
{
"epoch": 4.75,
"learning_rate": 3e-06,
"loss": 0.0034,
"step": 4525
},
{
"epoch": 4.77,
"learning_rate": 3e-06,
"loss": 0.0032,
"step": 4550
},
{
"epoch": 4.8,
"learning_rate": 3e-06,
"loss": 0.0041,
"step": 4575
},
{
"epoch": 4.83,
"learning_rate": 3e-06,
"loss": 0.0044,
"step": 4600
},
{
"epoch": 4.85,
"learning_rate": 3e-06,
"loss": 0.004,
"step": 4625
},
{
"epoch": 4.88,
"learning_rate": 3e-06,
"loss": 0.0041,
"step": 4650
},
{
"epoch": 4.91,
"learning_rate": 3e-06,
"loss": 0.0036,
"step": 4675
},
{
"epoch": 4.93,
"learning_rate": 3e-06,
"loss": 0.0031,
"step": 4700
},
{
"epoch": 4.96,
"learning_rate": 3e-06,
"loss": 0.0034,
"step": 4725
},
{
"epoch": 4.98,
"learning_rate": 3e-06,
"loss": 0.0034,
"step": 4750
},
{
"epoch": 5.01,
"learning_rate": 3e-06,
"loss": 0.0023,
"step": 4775
},
{
"epoch": 5.04,
"learning_rate": 3e-06,
"loss": 0.0026,
"step": 4800
},
{
"epoch": 5.06,
"learning_rate": 3e-06,
"loss": 0.0018,
"step": 4825
},
{
"epoch": 5.09,
"learning_rate": 3e-06,
"loss": 0.0016,
"step": 4850
},
{
"epoch": 5.12,
"learning_rate": 3e-06,
"loss": 0.0025,
"step": 4875
},
{
"epoch": 5.14,
"learning_rate": 3e-06,
"loss": 0.0018,
"step": 4900
},
{
"epoch": 5.17,
"learning_rate": 3e-06,
"loss": 0.0017,
"step": 4925
},
{
"epoch": 5.19,
"learning_rate": 3e-06,
"loss": 0.0021,
"step": 4950
},
{
"epoch": 5.22,
"learning_rate": 3e-06,
"loss": 0.0017,
"step": 4975
},
{
"epoch": 5.25,
"learning_rate": 3e-06,
"loss": 0.0016,
"step": 5000
},
{
"epoch": 5.25,
"eval_loss": 0.323974609375,
"eval_runtime": 824.2399,
"eval_samples_per_second": 5.619,
"eval_steps_per_second": 0.352,
"eval_wer": 18.965312777942483,
"step": 5000
},
{
"epoch": 5.27,
"learning_rate": 3e-06,
"loss": 0.0015,
"step": 5025
},
{
"epoch": 5.3,
"learning_rate": 3e-06,
"loss": 0.0016,
"step": 5050
},
{
"epoch": 5.33,
"learning_rate": 3e-06,
"loss": 0.0017,
"step": 5075
},
{
"epoch": 5.35,
"learning_rate": 3e-06,
"loss": 0.002,
"step": 5100
},
{
"epoch": 5.38,
"learning_rate": 3e-06,
"loss": 0.0014,
"step": 5125
},
{
"epoch": 5.4,
"learning_rate": 3e-06,
"loss": 0.0022,
"step": 5150
},
{
"epoch": 5.43,
"learning_rate": 3e-06,
"loss": 0.0018,
"step": 5175
},
{
"epoch": 5.46,
"learning_rate": 3e-06,
"loss": 0.0022,
"step": 5200
},
{
"epoch": 5.48,
"learning_rate": 3e-06,
"loss": 0.002,
"step": 5225
},
{
"epoch": 5.51,
"learning_rate": 3e-06,
"loss": 0.0022,
"step": 5250
},
{
"epoch": 5.54,
"learning_rate": 3e-06,
"loss": 0.0023,
"step": 5275
},
{
"epoch": 5.56,
"learning_rate": 3e-06,
"loss": 0.0022,
"step": 5300
},
{
"epoch": 5.59,
"learning_rate": 3e-06,
"loss": 0.0028,
"step": 5325
},
{
"epoch": 5.61,
"learning_rate": 3e-06,
"loss": 0.0025,
"step": 5350
},
{
"epoch": 5.64,
"learning_rate": 3e-06,
"loss": 0.0028,
"step": 5375
},
{
"epoch": 5.67,
"learning_rate": 3e-06,
"loss": 0.0033,
"step": 5400
},
{
"epoch": 5.69,
"learning_rate": 3e-06,
"loss": 0.003,
"step": 5425
},
{
"epoch": 5.72,
"learning_rate": 3e-06,
"loss": 0.0032,
"step": 5450
},
{
"epoch": 5.75,
"learning_rate": 3e-06,
"loss": 0.0036,
"step": 5475
},
{
"epoch": 5.77,
"learning_rate": 3e-06,
"loss": 0.0029,
"step": 5500
},
{
"epoch": 5.8,
"learning_rate": 3e-06,
"loss": 0.0033,
"step": 5525
},
{
"epoch": 5.82,
"learning_rate": 3e-06,
"loss": 0.0039,
"step": 5550
},
{
"epoch": 5.85,
"learning_rate": 3e-06,
"loss": 0.0029,
"step": 5575
},
{
"epoch": 5.88,
"learning_rate": 3e-06,
"loss": 0.0029,
"step": 5600
},
{
"epoch": 5.9,
"learning_rate": 3e-06,
"loss": 0.0019,
"step": 5625
},
{
"epoch": 5.93,
"learning_rate": 3e-06,
"loss": 0.0027,
"step": 5650
},
{
"epoch": 5.95,
"learning_rate": 3e-06,
"loss": 0.003,
"step": 5675
},
{
"epoch": 5.98,
"learning_rate": 3e-06,
"loss": 0.0027,
"step": 5700
},
{
"epoch": 6.01,
"learning_rate": 3e-06,
"loss": 0.002,
"step": 5725
},
{
"epoch": 6.03,
"learning_rate": 3e-06,
"loss": 0.002,
"step": 5750
},
{
"epoch": 6.06,
"learning_rate": 3e-06,
"loss": 0.0017,
"step": 5775
},
{
"epoch": 6.09,
"learning_rate": 3e-06,
"loss": 0.0017,
"step": 5800
},
{
"epoch": 6.11,
"learning_rate": 3e-06,
"loss": 0.002,
"step": 5825
},
{
"epoch": 6.14,
"learning_rate": 3e-06,
"loss": 0.002,
"step": 5850
},
{
"epoch": 6.16,
"learning_rate": 3e-06,
"loss": 0.003,
"step": 5875
},
{
"epoch": 6.19,
"learning_rate": 3e-06,
"loss": 0.0016,
"step": 5900
},
{
"epoch": 6.22,
"learning_rate": 3e-06,
"loss": 0.0017,
"step": 5925
},
{
"epoch": 6.24,
"learning_rate": 3e-06,
"loss": 0.0017,
"step": 5950
},
{
"epoch": 6.27,
"learning_rate": 3e-06,
"loss": 0.0015,
"step": 5975
},
{
"epoch": 6.3,
"learning_rate": 3e-06,
"loss": 0.0018,
"step": 6000
},
{
"epoch": 6.3,
"eval_loss": 0.331298828125,
"eval_runtime": 820.3708,
"eval_samples_per_second": 5.645,
"eval_steps_per_second": 0.353,
"eval_wer": 18.64512303587311,
"step": 6000
},
{
"epoch": 6.32,
"learning_rate": 3e-06,
"loss": 0.0018,
"step": 6025
},
{
"epoch": 6.35,
"learning_rate": 3e-06,
"loss": 0.0021,
"step": 6050
},
{
"epoch": 6.37,
"learning_rate": 3e-06,
"loss": 0.0014,
"step": 6075
},
{
"epoch": 6.4,
"learning_rate": 3e-06,
"loss": 0.0019,
"step": 6100
},
{
"epoch": 6.43,
"learning_rate": 3e-06,
"loss": 0.0014,
"step": 6125
},
{
"epoch": 6.45,
"learning_rate": 3e-06,
"loss": 0.0016,
"step": 6150
},
{
"epoch": 6.48,
"learning_rate": 3e-06,
"loss": 0.0022,
"step": 6175
},
{
"epoch": 6.51,
"learning_rate": 3e-06,
"loss": 0.0019,
"step": 6200
},
{
"epoch": 6.53,
"learning_rate": 3e-06,
"loss": 0.0027,
"step": 6225
},
{
"epoch": 6.56,
"learning_rate": 3e-06,
"loss": 0.0021,
"step": 6250
},
{
"epoch": 6.58,
"learning_rate": 3e-06,
"loss": 0.002,
"step": 6275
},
{
"epoch": 6.61,
"learning_rate": 3e-06,
"loss": 0.0023,
"step": 6300
},
{
"epoch": 6.64,
"learning_rate": 3e-06,
"loss": 0.0027,
"step": 6325
},
{
"epoch": 6.66,
"learning_rate": 3e-06,
"loss": 0.0021,
"step": 6350
},
{
"epoch": 6.69,
"learning_rate": 3e-06,
"loss": 0.0016,
"step": 6375
},
{
"epoch": 6.72,
"learning_rate": 3e-06,
"loss": 0.0019,
"step": 6400
},
{
"epoch": 6.74,
"learning_rate": 3e-06,
"loss": 0.0025,
"step": 6425
},
{
"epoch": 6.77,
"learning_rate": 3e-06,
"loss": 0.0022,
"step": 6450
},
{
"epoch": 6.79,
"learning_rate": 3e-06,
"loss": 0.0018,
"step": 6475
},
{
"epoch": 6.82,
"learning_rate": 3e-06,
"loss": 0.0025,
"step": 6500
},
{
"epoch": 6.85,
"learning_rate": 3e-06,
"loss": 0.0026,
"step": 6525
},
{
"epoch": 6.87,
"learning_rate": 3e-06,
"loss": 0.0028,
"step": 6550
},
{
"epoch": 6.9,
"learning_rate": 3e-06,
"loss": 0.0023,
"step": 6575
},
{
"epoch": 6.93,
"learning_rate": 3e-06,
"loss": 0.0028,
"step": 6600
},
{
"epoch": 6.95,
"learning_rate": 3e-06,
"loss": 0.0034,
"step": 6625
},
{
"epoch": 6.98,
"learning_rate": 3e-06,
"loss": 0.002,
"step": 6650
},
{
"epoch": 7.0,
"learning_rate": 3e-06,
"loss": 0.0022,
"step": 6675
},
{
"epoch": 7.03,
"learning_rate": 3e-06,
"loss": 0.0017,
"step": 6700
},
{
"epoch": 7.06,
"learning_rate": 3e-06,
"loss": 0.0018,
"step": 6725
},
{
"epoch": 7.08,
"learning_rate": 3e-06,
"loss": 0.0021,
"step": 6750
},
{
"epoch": 7.11,
"learning_rate": 3e-06,
"loss": 0.0019,
"step": 6775
},
{
"epoch": 7.14,
"learning_rate": 3e-06,
"loss": 0.0012,
"step": 6800
},
{
"epoch": 7.16,
"learning_rate": 3e-06,
"loss": 0.0013,
"step": 6825
},
{
"epoch": 7.19,
"learning_rate": 3e-06,
"loss": 0.0016,
"step": 6850
},
{
"epoch": 7.21,
"learning_rate": 3e-06,
"loss": 0.0018,
"step": 6875
},
{
"epoch": 7.24,
"learning_rate": 3e-06,
"loss": 0.0018,
"step": 6900
},
{
"epoch": 7.27,
"learning_rate": 3e-06,
"loss": 0.003,
"step": 6925
},
{
"epoch": 7.29,
"learning_rate": 3e-06,
"loss": 0.0016,
"step": 6950
},
{
"epoch": 7.32,
"learning_rate": 3e-06,
"loss": 0.0019,
"step": 6975
},
{
"epoch": 7.35,
"learning_rate": 3e-06,
"loss": 0.0014,
"step": 7000
},
{
"epoch": 7.35,
"eval_loss": 0.3330078125,
"eval_runtime": 820.6102,
"eval_samples_per_second": 5.643,
"eval_steps_per_second": 0.353,
"eval_wer": 18.944559739104655,
"step": 7000
},
{
"epoch": 7.37,
"learning_rate": 3e-06,
"loss": 0.0023,
"step": 7025
},
{
"epoch": 7.4,
"learning_rate": 3e-06,
"loss": 0.0016,
"step": 7050
},
{
"epoch": 7.42,
"learning_rate": 3e-06,
"loss": 0.0018,
"step": 7075
},
{
"epoch": 7.45,
"learning_rate": 3e-06,
"loss": 0.0017,
"step": 7100
},
{
"epoch": 7.48,
"learning_rate": 3e-06,
"loss": 0.0018,
"step": 7125
},
{
"epoch": 7.5,
"learning_rate": 3e-06,
"loss": 0.0016,
"step": 7150
},
{
"epoch": 7.53,
"learning_rate": 3e-06,
"loss": 0.0023,
"step": 7175
},
{
"epoch": 7.56,
"learning_rate": 3e-06,
"loss": 0.0022,
"step": 7200
},
{
"epoch": 7.58,
"learning_rate": 3e-06,
"loss": 0.0024,
"step": 7225
},
{
"epoch": 7.61,
"learning_rate": 3e-06,
"loss": 0.003,
"step": 7250
},
{
"epoch": 7.63,
"learning_rate": 3e-06,
"loss": 0.0019,
"step": 7275
},
{
"epoch": 7.66,
"learning_rate": 3e-06,
"loss": 0.0025,
"step": 7300
},
{
"epoch": 7.69,
"learning_rate": 3e-06,
"loss": 0.0028,
"step": 7325
},
{
"epoch": 7.71,
"learning_rate": 3e-06,
"loss": 0.0016,
"step": 7350
},
{
"epoch": 7.74,
"learning_rate": 3e-06,
"loss": 0.0023,
"step": 7375
},
{
"epoch": 7.76,
"learning_rate": 3e-06,
"loss": 0.0021,
"step": 7400
},
{
"epoch": 7.79,
"learning_rate": 3e-06,
"loss": 0.0021,
"step": 7425
},
{
"epoch": 7.82,
"learning_rate": 3e-06,
"loss": 0.0026,
"step": 7450
},
{
"epoch": 7.84,
"learning_rate": 3e-06,
"loss": 0.0016,
"step": 7475
},
{
"epoch": 7.87,
"learning_rate": 3e-06,
"loss": 0.0019,
"step": 7500
},
{
"epoch": 7.9,
"learning_rate": 3e-06,
"loss": 0.0022,
"step": 7525
},
{
"epoch": 7.92,
"learning_rate": 3e-06,
"loss": 0.0025,
"step": 7550
},
{
"epoch": 7.95,
"learning_rate": 3e-06,
"loss": 0.0023,
"step": 7575
},
{
"epoch": 7.97,
"learning_rate": 3e-06,
"loss": 0.0033,
"step": 7600
},
{
"epoch": 8.0,
"learning_rate": 3e-06,
"loss": 0.0028,
"step": 7625
},
{
"epoch": 8.03,
"learning_rate": 3e-06,
"loss": 0.0015,
"step": 7650
},
{
"epoch": 8.05,
"learning_rate": 3e-06,
"loss": 0.0013,
"step": 7675
},
{
"epoch": 8.08,
"learning_rate": 3e-06,
"loss": 0.0013,
"step": 7700
},
{
"epoch": 8.11,
"learning_rate": 3e-06,
"loss": 0.0015,
"step": 7725
},
{
"epoch": 8.13,
"learning_rate": 3e-06,
"loss": 0.0016,
"step": 7750
},
{
"epoch": 8.16,
"learning_rate": 3e-06,
"loss": 0.0017,
"step": 7775
},
{
"epoch": 8.18,
"learning_rate": 3e-06,
"loss": 0.0014,
"step": 7800
},
{
"epoch": 8.21,
"learning_rate": 3e-06,
"loss": 0.0015,
"step": 7825
},
{
"epoch": 8.24,
"learning_rate": 3e-06,
"loss": 0.0013,
"step": 7850
},
{
"epoch": 8.26,
"learning_rate": 3e-06,
"loss": 0.0015,
"step": 7875
},
{
"epoch": 8.29,
"learning_rate": 3e-06,
"loss": 0.0013,
"step": 7900
},
{
"epoch": 8.32,
"learning_rate": 3e-06,
"loss": 0.0011,
"step": 7925
},
{
"epoch": 8.34,
"learning_rate": 3e-06,
"loss": 0.0017,
"step": 7950
},
{
"epoch": 8.37,
"learning_rate": 3e-06,
"loss": 0.0015,
"step": 7975
},
{
"epoch": 8.39,
"learning_rate": 3e-06,
"loss": 0.0016,
"step": 8000
},
{
"epoch": 8.39,
"eval_loss": 0.3427734375,
"eval_runtime": 817.7486,
"eval_samples_per_second": 5.663,
"eval_steps_per_second": 0.355,
"eval_wer": 18.642158316039133,
"step": 8000
},
{
"epoch": 8.42,
"learning_rate": 3e-06,
"loss": 0.0012,
"step": 8025
},
{
"epoch": 8.45,
"learning_rate": 3e-06,
"loss": 0.0018,
"step": 8050
},
{
"epoch": 8.47,
"learning_rate": 3e-06,
"loss": 0.0012,
"step": 8075
},
{
"epoch": 8.5,
"learning_rate": 3e-06,
"loss": 0.0021,
"step": 8100
},
{
"epoch": 8.53,
"learning_rate": 3e-06,
"loss": 0.0012,
"step": 8125
},
{
"epoch": 8.55,
"learning_rate": 3e-06,
"loss": 0.0015,
"step": 8150
},
{
"epoch": 8.58,
"learning_rate": 3e-06,
"loss": 0.0014,
"step": 8175
},
{
"epoch": 8.6,
"learning_rate": 3e-06,
"loss": 0.0016,
"step": 8200
},
{
"epoch": 8.63,
"learning_rate": 3e-06,
"loss": 0.002,
"step": 8225
},
{
"epoch": 8.66,
"learning_rate": 3e-06,
"loss": 0.0012,
"step": 8250
},
{
"epoch": 8.68,
"learning_rate": 3e-06,
"loss": 0.0012,
"step": 8275
},
{
"epoch": 8.71,
"learning_rate": 3e-06,
"loss": 0.0016,
"step": 8300
},
{
"epoch": 8.74,
"learning_rate": 3e-06,
"loss": 0.0012,
"step": 8325
},
{
"epoch": 8.76,
"learning_rate": 3e-06,
"loss": 0.0013,
"step": 8350
},
{
"epoch": 8.79,
"learning_rate": 3e-06,
"loss": 0.0012,
"step": 8375
},
{
"epoch": 8.81,
"learning_rate": 3e-06,
"loss": 0.0021,
"step": 8400
},
{
"epoch": 8.84,
"learning_rate": 3e-06,
"loss": 0.0018,
"step": 8425
},
{
"epoch": 8.87,
"learning_rate": 3e-06,
"loss": 0.0014,
"step": 8450
},
{
"epoch": 8.89,
"learning_rate": 3e-06,
"loss": 0.0031,
"step": 8475
},
{
"epoch": 8.92,
"learning_rate": 3e-06,
"loss": 0.0022,
"step": 8500
},
{
"epoch": 8.95,
"learning_rate": 3e-06,
"loss": 0.0026,
"step": 8525
},
{
"epoch": 8.97,
"learning_rate": 3e-06,
"loss": 0.0021,
"step": 8550
},
{
"epoch": 9.0,
"learning_rate": 3e-06,
"loss": 0.0024,
"step": 8575
},
{
"epoch": 9.02,
"learning_rate": 3e-06,
"loss": 0.0016,
"step": 8600
},
{
"epoch": 9.05,
"learning_rate": 3e-06,
"loss": 0.0021,
"step": 8625
},
{
"epoch": 9.08,
"learning_rate": 3e-06,
"loss": 0.0012,
"step": 8650
},
{
"epoch": 9.1,
"learning_rate": 3e-06,
"loss": 0.0016,
"step": 8675
},
{
"epoch": 9.13,
"learning_rate": 3e-06,
"loss": 0.0011,
"step": 8700
},
{
"epoch": 9.16,
"learning_rate": 3e-06,
"loss": 0.0012,
"step": 8725
},
{
"epoch": 9.18,
"learning_rate": 3e-06,
"loss": 0.0017,
"step": 8750
},
{
"epoch": 9.21,
"learning_rate": 3e-06,
"loss": 0.0019,
"step": 8775
},
{
"epoch": 9.23,
"learning_rate": 3e-06,
"loss": 0.0016,
"step": 8800
},
{
"epoch": 9.26,
"learning_rate": 3e-06,
"loss": 0.0018,
"step": 8825
},
{
"epoch": 9.29,
"learning_rate": 3e-06,
"loss": 0.0023,
"step": 8850
},
{
"epoch": 9.31,
"learning_rate": 3e-06,
"loss": 0.0025,
"step": 8875
},
{
"epoch": 9.34,
"learning_rate": 3e-06,
"loss": 0.0019,
"step": 8900
},
{
"epoch": 9.37,
"learning_rate": 3e-06,
"loss": 0.0016,
"step": 8925
},
{
"epoch": 9.39,
"learning_rate": 3e-06,
"loss": 0.002,
"step": 8950
},
{
"epoch": 9.42,
"learning_rate": 3e-06,
"loss": 0.0012,
"step": 8975
},
{
"epoch": 9.44,
"learning_rate": 3e-06,
"loss": 0.0015,
"step": 9000
},
{
"epoch": 9.44,
"eval_loss": 0.350830078125,
"eval_runtime": 821.3387,
"eval_samples_per_second": 5.638,
"eval_steps_per_second": 0.353,
"eval_wer": 18.95641861844056,
"step": 9000
},
{
"epoch": 9.47,
"learning_rate": 3e-06,
"loss": 0.0018,
"step": 9025
},
{
"epoch": 9.5,
"learning_rate": 3e-06,
"loss": 0.0008,
"step": 9050
},
{
"epoch": 9.52,
"learning_rate": 3e-06,
"loss": 0.0011,
"step": 9075
},
{
"epoch": 9.55,
"learning_rate": 3e-06,
"loss": 0.0021,
"step": 9100
},
{
"epoch": 9.58,
"learning_rate": 3e-06,
"loss": 0.0015,
"step": 9125
},
{
"epoch": 9.6,
"learning_rate": 3e-06,
"loss": 0.0021,
"step": 9150
},
{
"epoch": 9.63,
"learning_rate": 3e-06,
"loss": 0.0017,
"step": 9175
},
{
"epoch": 9.65,
"learning_rate": 3e-06,
"loss": 0.0009,
"step": 9200
},
{
"epoch": 9.68,
"learning_rate": 3e-06,
"loss": 0.0013,
"step": 9225
},
{
"epoch": 9.71,
"learning_rate": 3e-06,
"loss": 0.0008,
"step": 9250
},
{
"epoch": 9.73,
"learning_rate": 3e-06,
"loss": 0.0011,
"step": 9275
},
{
"epoch": 9.76,
"learning_rate": 3e-06,
"loss": 0.0008,
"step": 9300
},
{
"epoch": 9.78,
"learning_rate": 3e-06,
"loss": 0.0012,
"step": 9325
},
{
"epoch": 9.81,
"learning_rate": 3e-06,
"loss": 0.0015,
"step": 9350
},
{
"epoch": 9.84,
"learning_rate": 3e-06,
"loss": 0.0011,
"step": 9375
},
{
"epoch": 9.86,
"learning_rate": 3e-06,
"loss": 0.001,
"step": 9400
},
{
"epoch": 9.89,
"learning_rate": 3e-06,
"loss": 0.0008,
"step": 9425
},
{
"epoch": 9.92,
"learning_rate": 3e-06,
"loss": 0.002,
"step": 9450
},
{
"epoch": 9.94,
"learning_rate": 3e-06,
"loss": 0.002,
"step": 9475
},
{
"epoch": 9.97,
"learning_rate": 3e-06,
"loss": 0.0018,
"step": 9500
},
{
"epoch": 9.99,
"learning_rate": 3e-06,
"loss": 0.0019,
"step": 9525
},
{
"epoch": 10.02,
"learning_rate": 3e-06,
"loss": 0.0017,
"step": 9550
},
{
"epoch": 10.05,
"learning_rate": 3e-06,
"loss": 0.0015,
"step": 9575
},
{
"epoch": 10.07,
"learning_rate": 3e-06,
"loss": 0.0018,
"step": 9600
},
{
"epoch": 10.1,
"learning_rate": 3e-06,
"loss": 0.0018,
"step": 9625
},
{
"epoch": 10.13,
"learning_rate": 3e-06,
"loss": 0.0015,
"step": 9650
},
{
"epoch": 10.15,
"learning_rate": 3e-06,
"loss": 0.0026,
"step": 9675
},
{
"epoch": 10.18,
"learning_rate": 3e-06,
"loss": 0.0013,
"step": 9700
},
{
"epoch": 10.2,
"learning_rate": 3e-06,
"loss": 0.0011,
"step": 9725
},
{
"epoch": 10.23,
"learning_rate": 3e-06,
"loss": 0.0018,
"step": 9750
},
{
"epoch": 10.26,
"learning_rate": 3e-06,
"loss": 0.0013,
"step": 9775
},
{
"epoch": 10.28,
"learning_rate": 3e-06,
"loss": 0.0016,
"step": 9800
},
{
"epoch": 10.31,
"learning_rate": 3e-06,
"loss": 0.0017,
"step": 9825
},
{
"epoch": 10.34,
"learning_rate": 3e-06,
"loss": 0.0012,
"step": 9850
},
{
"epoch": 10.36,
"learning_rate": 3e-06,
"loss": 0.0017,
"step": 9875
},
{
"epoch": 10.39,
"learning_rate": 3e-06,
"loss": 0.0009,
"step": 9900
},
{
"epoch": 10.41,
"learning_rate": 3e-06,
"loss": 0.0011,
"step": 9925
},
{
"epoch": 10.44,
"learning_rate": 3e-06,
"loss": 0.001,
"step": 9950
},
{
"epoch": 10.47,
"learning_rate": 3e-06,
"loss": 0.0012,
"step": 9975
},
{
"epoch": 10.49,
"learning_rate": 3e-06,
"loss": 0.001,
"step": 10000
},
{
"epoch": 10.49,
"eval_loss": 0.35693359375,
"eval_runtime": 821.9897,
"eval_samples_per_second": 5.634,
"eval_steps_per_second": 0.353,
"eval_wer": 18.855618144085383,
"step": 10000
},
{
"epoch": 10.49,
"step": 10000,
"total_flos": 3.265629558325379e+20,
"train_loss": 0.026331889963150026,
"train_runtime": 46193.1062,
"train_samples_per_second": 6.927,
"train_steps_per_second": 0.216
}
],
"max_steps": 10000,
"num_train_epochs": 11,
"total_flos": 3.265629558325379e+20,
"trial_name": null,
"trial_params": null
}