whisper-small-bn-crblp / trainer_state.json
Rakib's picture
End of training
aca7f7d
raw
history blame
179 kB
{
"best_metric": 9.653842723732433,
"best_model_checkpoint": "./checkpoint-25000",
"epoch": 4.472772000447277,
"global_step": 35000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 3.1250000000000005e-08,
"loss": 2.803,
"step": 25
},
{
"epoch": 0.01,
"learning_rate": 6.250000000000001e-08,
"loss": 2.7814,
"step": 50
},
{
"epoch": 0.01,
"learning_rate": 9.375e-08,
"loss": 2.7507,
"step": 75
},
{
"epoch": 0.01,
"learning_rate": 1.2500000000000002e-07,
"loss": 2.7002,
"step": 100
},
{
"epoch": 0.02,
"learning_rate": 1.5625e-07,
"loss": 2.5888,
"step": 125
},
{
"epoch": 0.02,
"learning_rate": 1.875e-07,
"loss": 2.4547,
"step": 150
},
{
"epoch": 0.02,
"learning_rate": 2.1875e-07,
"loss": 2.2567,
"step": 175
},
{
"epoch": 0.03,
"learning_rate": 2.5000000000000004e-07,
"loss": 2.084,
"step": 200
},
{
"epoch": 0.03,
"learning_rate": 2.8125e-07,
"loss": 1.895,
"step": 225
},
{
"epoch": 0.03,
"learning_rate": 3.125e-07,
"loss": 1.7252,
"step": 250
},
{
"epoch": 0.04,
"learning_rate": 3.4375000000000004e-07,
"loss": 1.5906,
"step": 275
},
{
"epoch": 0.04,
"learning_rate": 3.75e-07,
"loss": 1.4839,
"step": 300
},
{
"epoch": 0.04,
"learning_rate": 4.0625000000000003e-07,
"loss": 1.3814,
"step": 325
},
{
"epoch": 0.04,
"learning_rate": 4.375e-07,
"loss": 1.2917,
"step": 350
},
{
"epoch": 0.05,
"learning_rate": 4.6875000000000006e-07,
"loss": 1.2315,
"step": 375
},
{
"epoch": 0.05,
"learning_rate": 5.000000000000001e-07,
"loss": 1.1606,
"step": 400
},
{
"epoch": 0.05,
"learning_rate": 5.3125e-07,
"loss": 1.0859,
"step": 425
},
{
"epoch": 0.06,
"learning_rate": 5.625e-07,
"loss": 1.0429,
"step": 450
},
{
"epoch": 0.06,
"learning_rate": 5.9375e-07,
"loss": 0.9984,
"step": 475
},
{
"epoch": 0.06,
"learning_rate": 6.25e-07,
"loss": 0.9521,
"step": 500
},
{
"epoch": 0.07,
"learning_rate": 6.562500000000001e-07,
"loss": 0.9111,
"step": 525
},
{
"epoch": 0.07,
"learning_rate": 6.875000000000001e-07,
"loss": 0.883,
"step": 550
},
{
"epoch": 0.07,
"learning_rate": 7.1875e-07,
"loss": 0.8423,
"step": 575
},
{
"epoch": 0.08,
"learning_rate": 7.5e-07,
"loss": 0.8191,
"step": 600
},
{
"epoch": 0.08,
"learning_rate": 7.8125e-07,
"loss": 0.788,
"step": 625
},
{
"epoch": 0.08,
"learning_rate": 8.125000000000001e-07,
"loss": 0.7726,
"step": 650
},
{
"epoch": 0.09,
"learning_rate": 8.437500000000001e-07,
"loss": 0.7423,
"step": 675
},
{
"epoch": 0.09,
"learning_rate": 8.75e-07,
"loss": 0.7225,
"step": 700
},
{
"epoch": 0.09,
"learning_rate": 9.0625e-07,
"loss": 0.7073,
"step": 725
},
{
"epoch": 0.1,
"learning_rate": 9.375000000000001e-07,
"loss": 0.683,
"step": 750
},
{
"epoch": 0.1,
"learning_rate": 9.6875e-07,
"loss": 0.6654,
"step": 775
},
{
"epoch": 0.1,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.6566,
"step": 800
},
{
"epoch": 0.11,
"learning_rate": 1.03125e-06,
"loss": 0.6476,
"step": 825
},
{
"epoch": 0.11,
"learning_rate": 1.0625e-06,
"loss": 0.6158,
"step": 850
},
{
"epoch": 0.11,
"learning_rate": 1.0937500000000001e-06,
"loss": 0.6013,
"step": 875
},
{
"epoch": 0.12,
"learning_rate": 1.125e-06,
"loss": 0.5909,
"step": 900
},
{
"epoch": 0.12,
"learning_rate": 1.1562500000000002e-06,
"loss": 0.5771,
"step": 925
},
{
"epoch": 0.12,
"learning_rate": 1.1875e-06,
"loss": 0.559,
"step": 950
},
{
"epoch": 0.12,
"learning_rate": 1.21875e-06,
"loss": 0.5536,
"step": 975
},
{
"epoch": 0.13,
"learning_rate": 1.25e-06,
"loss": 0.5361,
"step": 1000
},
{
"epoch": 0.13,
"eval_cer": 22.659873675496076,
"eval_loss": 0.40432098507881165,
"eval_runtime": 3456.2826,
"eval_samples_per_second": 2.631,
"eval_steps_per_second": 0.082,
"eval_wer": 44.05213380840953,
"step": 1000
},
{
"epoch": 0.13,
"learning_rate": 1.28125e-06,
"loss": 0.5145,
"step": 1025
},
{
"epoch": 0.13,
"learning_rate": 1.3125000000000001e-06,
"loss": 0.5056,
"step": 1050
},
{
"epoch": 0.14,
"learning_rate": 1.34375e-06,
"loss": 0.4904,
"step": 1075
},
{
"epoch": 0.14,
"learning_rate": 1.3750000000000002e-06,
"loss": 0.4781,
"step": 1100
},
{
"epoch": 0.14,
"learning_rate": 1.40625e-06,
"loss": 0.4463,
"step": 1125
},
{
"epoch": 0.15,
"learning_rate": 1.4375e-06,
"loss": 0.3987,
"step": 1150
},
{
"epoch": 0.15,
"learning_rate": 1.4687500000000001e-06,
"loss": 0.3967,
"step": 1175
},
{
"epoch": 0.15,
"learning_rate": 1.5e-06,
"loss": 0.3805,
"step": 1200
},
{
"epoch": 0.16,
"learning_rate": 1.5312500000000002e-06,
"loss": 0.383,
"step": 1225
},
{
"epoch": 0.16,
"learning_rate": 1.5625e-06,
"loss": 0.3788,
"step": 1250
},
{
"epoch": 0.16,
"learning_rate": 1.59375e-06,
"loss": 0.3614,
"step": 1275
},
{
"epoch": 0.17,
"learning_rate": 1.6250000000000001e-06,
"loss": 0.3592,
"step": 1300
},
{
"epoch": 0.17,
"learning_rate": 1.65625e-06,
"loss": 0.3527,
"step": 1325
},
{
"epoch": 0.17,
"learning_rate": 1.6875000000000001e-06,
"loss": 0.3436,
"step": 1350
},
{
"epoch": 0.18,
"learning_rate": 1.71875e-06,
"loss": 0.3316,
"step": 1375
},
{
"epoch": 0.18,
"learning_rate": 1.75e-06,
"loss": 0.3531,
"step": 1400
},
{
"epoch": 0.18,
"learning_rate": 1.78125e-06,
"loss": 0.3467,
"step": 1425
},
{
"epoch": 0.19,
"learning_rate": 1.8125e-06,
"loss": 0.3351,
"step": 1450
},
{
"epoch": 0.19,
"learning_rate": 1.8437500000000003e-06,
"loss": 0.3467,
"step": 1475
},
{
"epoch": 0.19,
"learning_rate": 1.8750000000000003e-06,
"loss": 0.3275,
"step": 1500
},
{
"epoch": 0.19,
"learning_rate": 1.90625e-06,
"loss": 0.3352,
"step": 1525
},
{
"epoch": 0.2,
"learning_rate": 1.9375e-06,
"loss": 0.3281,
"step": 1550
},
{
"epoch": 0.2,
"learning_rate": 1.96875e-06,
"loss": 0.3206,
"step": 1575
},
{
"epoch": 0.2,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.3142,
"step": 1600
},
{
"epoch": 0.21,
"learning_rate": 2.0312500000000002e-06,
"loss": 0.3156,
"step": 1625
},
{
"epoch": 0.21,
"learning_rate": 2.0625e-06,
"loss": 0.2933,
"step": 1650
},
{
"epoch": 0.21,
"learning_rate": 2.09375e-06,
"loss": 0.3099,
"step": 1675
},
{
"epoch": 0.22,
"learning_rate": 2.125e-06,
"loss": 0.3092,
"step": 1700
},
{
"epoch": 0.22,
"learning_rate": 2.1562500000000003e-06,
"loss": 0.3011,
"step": 1725
},
{
"epoch": 0.22,
"learning_rate": 2.1875000000000002e-06,
"loss": 0.3019,
"step": 1750
},
{
"epoch": 0.23,
"learning_rate": 2.21875e-06,
"loss": 0.2926,
"step": 1775
},
{
"epoch": 0.23,
"learning_rate": 2.25e-06,
"loss": 0.2961,
"step": 1800
},
{
"epoch": 0.23,
"learning_rate": 2.28125e-06,
"loss": 0.29,
"step": 1825
},
{
"epoch": 0.24,
"learning_rate": 2.3125000000000003e-06,
"loss": 0.2974,
"step": 1850
},
{
"epoch": 0.24,
"learning_rate": 2.3437500000000002e-06,
"loss": 0.3033,
"step": 1875
},
{
"epoch": 0.24,
"learning_rate": 2.375e-06,
"loss": 0.2882,
"step": 1900
},
{
"epoch": 0.25,
"learning_rate": 2.40625e-06,
"loss": 0.2887,
"step": 1925
},
{
"epoch": 0.25,
"learning_rate": 2.4375e-06,
"loss": 0.2845,
"step": 1950
},
{
"epoch": 0.25,
"learning_rate": 2.4687500000000003e-06,
"loss": 0.2929,
"step": 1975
},
{
"epoch": 0.26,
"learning_rate": 2.5e-06,
"loss": 0.2881,
"step": 2000
},
{
"epoch": 0.26,
"eval_cer": 16.393912178968336,
"eval_loss": 0.22170716524124146,
"eval_runtime": 3526.7853,
"eval_samples_per_second": 2.578,
"eval_steps_per_second": 0.081,
"eval_wer": 32.48935113028781,
"step": 2000
},
{
"epoch": 0.26,
"learning_rate": 2.53125e-06,
"loss": 0.2766,
"step": 2025
},
{
"epoch": 0.26,
"learning_rate": 2.5625e-06,
"loss": 0.276,
"step": 2050
},
{
"epoch": 0.27,
"learning_rate": 2.5937500000000004e-06,
"loss": 0.265,
"step": 2075
},
{
"epoch": 0.27,
"learning_rate": 2.6250000000000003e-06,
"loss": 0.2781,
"step": 2100
},
{
"epoch": 0.27,
"learning_rate": 2.65625e-06,
"loss": 0.2778,
"step": 2125
},
{
"epoch": 0.27,
"learning_rate": 2.6875e-06,
"loss": 0.2673,
"step": 2150
},
{
"epoch": 0.28,
"learning_rate": 2.71875e-06,
"loss": 0.2652,
"step": 2175
},
{
"epoch": 0.28,
"learning_rate": 2.7500000000000004e-06,
"loss": 0.2552,
"step": 2200
},
{
"epoch": 0.28,
"learning_rate": 2.7812500000000003e-06,
"loss": 0.2617,
"step": 2225
},
{
"epoch": 0.29,
"learning_rate": 2.8125e-06,
"loss": 0.2593,
"step": 2250
},
{
"epoch": 0.29,
"learning_rate": 2.84375e-06,
"loss": 0.2592,
"step": 2275
},
{
"epoch": 0.29,
"learning_rate": 2.875e-06,
"loss": 0.2559,
"step": 2300
},
{
"epoch": 0.3,
"learning_rate": 2.9062500000000003e-06,
"loss": 0.2589,
"step": 2325
},
{
"epoch": 0.3,
"learning_rate": 2.9375000000000003e-06,
"loss": 0.2559,
"step": 2350
},
{
"epoch": 0.3,
"learning_rate": 2.96875e-06,
"loss": 0.2537,
"step": 2375
},
{
"epoch": 0.31,
"learning_rate": 3e-06,
"loss": 0.2442,
"step": 2400
},
{
"epoch": 0.31,
"learning_rate": 3.03125e-06,
"loss": 0.2535,
"step": 2425
},
{
"epoch": 0.31,
"learning_rate": 3.0625000000000003e-06,
"loss": 0.25,
"step": 2450
},
{
"epoch": 0.32,
"learning_rate": 3.0937500000000002e-06,
"loss": 0.2531,
"step": 2475
},
{
"epoch": 0.32,
"learning_rate": 3.125e-06,
"loss": 0.2541,
"step": 2500
},
{
"epoch": 0.32,
"learning_rate": 3.15625e-06,
"loss": 0.2501,
"step": 2525
},
{
"epoch": 0.33,
"learning_rate": 3.1875e-06,
"loss": 0.2408,
"step": 2550
},
{
"epoch": 0.33,
"learning_rate": 3.2187500000000003e-06,
"loss": 0.2365,
"step": 2575
},
{
"epoch": 0.33,
"learning_rate": 3.2500000000000002e-06,
"loss": 0.2409,
"step": 2600
},
{
"epoch": 0.34,
"learning_rate": 3.28125e-06,
"loss": 0.2401,
"step": 2625
},
{
"epoch": 0.34,
"learning_rate": 3.3125e-06,
"loss": 0.238,
"step": 2650
},
{
"epoch": 0.34,
"learning_rate": 3.34375e-06,
"loss": 0.238,
"step": 2675
},
{
"epoch": 0.35,
"learning_rate": 3.3750000000000003e-06,
"loss": 0.2241,
"step": 2700
},
{
"epoch": 0.35,
"learning_rate": 3.40625e-06,
"loss": 0.2396,
"step": 2725
},
{
"epoch": 0.35,
"learning_rate": 3.4375e-06,
"loss": 0.2295,
"step": 2750
},
{
"epoch": 0.35,
"learning_rate": 3.46875e-06,
"loss": 0.2369,
"step": 2775
},
{
"epoch": 0.36,
"learning_rate": 3.5e-06,
"loss": 0.2303,
"step": 2800
},
{
"epoch": 0.36,
"learning_rate": 3.5312500000000007e-06,
"loss": 0.2225,
"step": 2825
},
{
"epoch": 0.36,
"learning_rate": 3.5625e-06,
"loss": 0.2259,
"step": 2850
},
{
"epoch": 0.37,
"learning_rate": 3.59375e-06,
"loss": 0.2144,
"step": 2875
},
{
"epoch": 0.37,
"learning_rate": 3.625e-06,
"loss": 0.2297,
"step": 2900
},
{
"epoch": 0.37,
"learning_rate": 3.65625e-06,
"loss": 0.2268,
"step": 2925
},
{
"epoch": 0.38,
"learning_rate": 3.6875000000000007e-06,
"loss": 0.2208,
"step": 2950
},
{
"epoch": 0.38,
"learning_rate": 3.7187500000000006e-06,
"loss": 0.2229,
"step": 2975
},
{
"epoch": 0.38,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.2265,
"step": 3000
},
{
"epoch": 0.38,
"eval_cer": 13.04250608892979,
"eval_loss": 0.1728239357471466,
"eval_runtime": 3496.6554,
"eval_samples_per_second": 2.6,
"eval_steps_per_second": 0.082,
"eval_wer": 25.96370678342556,
"step": 3000
},
{
"epoch": 0.39,
"learning_rate": 3.78125e-06,
"loss": 0.2233,
"step": 3025
},
{
"epoch": 0.39,
"learning_rate": 3.8125e-06,
"loss": 0.2202,
"step": 3050
},
{
"epoch": 0.39,
"learning_rate": 3.84375e-06,
"loss": 0.2145,
"step": 3075
},
{
"epoch": 0.4,
"learning_rate": 3.875e-06,
"loss": 0.2188,
"step": 3100
},
{
"epoch": 0.4,
"learning_rate": 3.90625e-06,
"loss": 0.2135,
"step": 3125
},
{
"epoch": 0.4,
"learning_rate": 3.9375e-06,
"loss": 0.2209,
"step": 3150
},
{
"epoch": 0.41,
"learning_rate": 3.96875e-06,
"loss": 0.222,
"step": 3175
},
{
"epoch": 0.41,
"learning_rate": 4.000000000000001e-06,
"loss": 0.2132,
"step": 3200
},
{
"epoch": 0.41,
"learning_rate": 4.031250000000001e-06,
"loss": 0.2161,
"step": 3225
},
{
"epoch": 0.42,
"learning_rate": 4.0625000000000005e-06,
"loss": 0.215,
"step": 3250
},
{
"epoch": 0.42,
"learning_rate": 4.09375e-06,
"loss": 0.2147,
"step": 3275
},
{
"epoch": 0.42,
"learning_rate": 4.125e-06,
"loss": 0.2082,
"step": 3300
},
{
"epoch": 0.42,
"learning_rate": 4.15625e-06,
"loss": 0.2165,
"step": 3325
},
{
"epoch": 0.43,
"learning_rate": 4.1875e-06,
"loss": 0.2046,
"step": 3350
},
{
"epoch": 0.43,
"learning_rate": 4.21875e-06,
"loss": 0.2195,
"step": 3375
},
{
"epoch": 0.43,
"learning_rate": 4.25e-06,
"loss": 0.211,
"step": 3400
},
{
"epoch": 0.44,
"learning_rate": 4.28125e-06,
"loss": 0.195,
"step": 3425
},
{
"epoch": 0.44,
"learning_rate": 4.312500000000001e-06,
"loss": 0.2101,
"step": 3450
},
{
"epoch": 0.44,
"learning_rate": 4.3437500000000006e-06,
"loss": 0.2031,
"step": 3475
},
{
"epoch": 0.45,
"learning_rate": 4.3750000000000005e-06,
"loss": 0.2016,
"step": 3500
},
{
"epoch": 0.45,
"learning_rate": 4.40625e-06,
"loss": 0.2004,
"step": 3525
},
{
"epoch": 0.45,
"learning_rate": 4.4375e-06,
"loss": 0.2028,
"step": 3550
},
{
"epoch": 0.46,
"learning_rate": 4.46875e-06,
"loss": 0.2029,
"step": 3575
},
{
"epoch": 0.46,
"learning_rate": 4.5e-06,
"loss": 0.1969,
"step": 3600
},
{
"epoch": 0.46,
"learning_rate": 4.53125e-06,
"loss": 0.1936,
"step": 3625
},
{
"epoch": 0.47,
"learning_rate": 4.5625e-06,
"loss": 0.1872,
"step": 3650
},
{
"epoch": 0.47,
"learning_rate": 4.59375e-06,
"loss": 0.1959,
"step": 3675
},
{
"epoch": 0.47,
"learning_rate": 4.625000000000001e-06,
"loss": 0.2003,
"step": 3700
},
{
"epoch": 0.48,
"learning_rate": 4.6562500000000005e-06,
"loss": 0.1927,
"step": 3725
},
{
"epoch": 0.48,
"learning_rate": 4.6875000000000004e-06,
"loss": 0.1839,
"step": 3750
},
{
"epoch": 0.48,
"learning_rate": 4.71875e-06,
"loss": 0.1907,
"step": 3775
},
{
"epoch": 0.49,
"learning_rate": 4.75e-06,
"loss": 0.1908,
"step": 3800
},
{
"epoch": 0.49,
"learning_rate": 4.781250000000001e-06,
"loss": 0.1972,
"step": 3825
},
{
"epoch": 0.49,
"learning_rate": 4.8125e-06,
"loss": 0.1905,
"step": 3850
},
{
"epoch": 0.5,
"learning_rate": 4.84375e-06,
"loss": 0.1858,
"step": 3875
},
{
"epoch": 0.5,
"learning_rate": 4.875e-06,
"loss": 0.1838,
"step": 3900
},
{
"epoch": 0.5,
"learning_rate": 4.90625e-06,
"loss": 0.1947,
"step": 3925
},
{
"epoch": 0.5,
"learning_rate": 4.937500000000001e-06,
"loss": 0.1912,
"step": 3950
},
{
"epoch": 0.51,
"learning_rate": 4.9687500000000005e-06,
"loss": 0.1889,
"step": 3975
},
{
"epoch": 0.51,
"learning_rate": 5e-06,
"loss": 0.1974,
"step": 4000
},
{
"epoch": 0.51,
"eval_cer": 11.325963787961104,
"eval_loss": 0.14298802614212036,
"eval_runtime": 3496.7534,
"eval_samples_per_second": 2.6,
"eval_steps_per_second": 0.082,
"eval_wer": 22.31867193837033,
"step": 4000
},
{
"epoch": 0.51,
"learning_rate": 5.031250000000001e-06,
"loss": 0.1902,
"step": 4025
},
{
"epoch": 0.52,
"learning_rate": 5.0625e-06,
"loss": 0.1839,
"step": 4050
},
{
"epoch": 0.52,
"learning_rate": 5.093750000000001e-06,
"loss": 0.1924,
"step": 4075
},
{
"epoch": 0.52,
"learning_rate": 5.125e-06,
"loss": 0.183,
"step": 4100
},
{
"epoch": 0.53,
"learning_rate": 5.156250000000001e-06,
"loss": 0.1895,
"step": 4125
},
{
"epoch": 0.53,
"learning_rate": 5.187500000000001e-06,
"loss": 0.1834,
"step": 4150
},
{
"epoch": 0.53,
"learning_rate": 5.21875e-06,
"loss": 0.1798,
"step": 4175
},
{
"epoch": 0.54,
"learning_rate": 5.2500000000000006e-06,
"loss": 0.1712,
"step": 4200
},
{
"epoch": 0.54,
"learning_rate": 5.28125e-06,
"loss": 0.188,
"step": 4225
},
{
"epoch": 0.54,
"learning_rate": 5.3125e-06,
"loss": 0.1756,
"step": 4250
},
{
"epoch": 0.55,
"learning_rate": 5.343750000000001e-06,
"loss": 0.1742,
"step": 4275
},
{
"epoch": 0.55,
"learning_rate": 5.375e-06,
"loss": 0.1882,
"step": 4300
},
{
"epoch": 0.55,
"learning_rate": 5.406250000000001e-06,
"loss": 0.173,
"step": 4325
},
{
"epoch": 0.56,
"learning_rate": 5.4375e-06,
"loss": 0.1735,
"step": 4350
},
{
"epoch": 0.56,
"learning_rate": 5.468750000000001e-06,
"loss": 0.1709,
"step": 4375
},
{
"epoch": 0.56,
"learning_rate": 5.500000000000001e-06,
"loss": 0.1747,
"step": 4400
},
{
"epoch": 0.57,
"learning_rate": 5.531250000000001e-06,
"loss": 0.1771,
"step": 4425
},
{
"epoch": 0.57,
"learning_rate": 5.5625000000000005e-06,
"loss": 0.1829,
"step": 4450
},
{
"epoch": 0.57,
"learning_rate": 5.59375e-06,
"loss": 0.1676,
"step": 4475
},
{
"epoch": 0.58,
"learning_rate": 5.625e-06,
"loss": 0.1704,
"step": 4500
},
{
"epoch": 0.58,
"learning_rate": 5.656250000000001e-06,
"loss": 0.1687,
"step": 4525
},
{
"epoch": 0.58,
"learning_rate": 5.6875e-06,
"loss": 0.1772,
"step": 4550
},
{
"epoch": 0.58,
"learning_rate": 5.718750000000001e-06,
"loss": 0.1692,
"step": 4575
},
{
"epoch": 0.59,
"learning_rate": 5.75e-06,
"loss": 0.1789,
"step": 4600
},
{
"epoch": 0.59,
"learning_rate": 5.781250000000001e-06,
"loss": 0.1703,
"step": 4625
},
{
"epoch": 0.59,
"learning_rate": 5.812500000000001e-06,
"loss": 0.1754,
"step": 4650
},
{
"epoch": 0.6,
"learning_rate": 5.843750000000001e-06,
"loss": 0.1747,
"step": 4675
},
{
"epoch": 0.6,
"learning_rate": 5.8750000000000005e-06,
"loss": 0.1671,
"step": 4700
},
{
"epoch": 0.6,
"learning_rate": 5.90625e-06,
"loss": 0.1705,
"step": 4725
},
{
"epoch": 0.61,
"learning_rate": 5.9375e-06,
"loss": 0.1707,
"step": 4750
},
{
"epoch": 0.61,
"learning_rate": 5.968750000000001e-06,
"loss": 0.1733,
"step": 4775
},
{
"epoch": 0.61,
"learning_rate": 6e-06,
"loss": 0.1661,
"step": 4800
},
{
"epoch": 0.62,
"learning_rate": 6.031250000000001e-06,
"loss": 0.1703,
"step": 4825
},
{
"epoch": 0.62,
"learning_rate": 6.0625e-06,
"loss": 0.163,
"step": 4850
},
{
"epoch": 0.62,
"learning_rate": 6.093750000000001e-06,
"loss": 0.1704,
"step": 4875
},
{
"epoch": 0.63,
"learning_rate": 6.125000000000001e-06,
"loss": 0.1639,
"step": 4900
},
{
"epoch": 0.63,
"learning_rate": 6.1562500000000006e-06,
"loss": 0.1652,
"step": 4925
},
{
"epoch": 0.63,
"learning_rate": 6.1875000000000005e-06,
"loss": 0.1721,
"step": 4950
},
{
"epoch": 0.64,
"learning_rate": 6.21875e-06,
"loss": 0.1665,
"step": 4975
},
{
"epoch": 0.64,
"learning_rate": 6.25e-06,
"loss": 0.1591,
"step": 5000
},
{
"epoch": 0.64,
"eval_cer": 10.016705288406456,
"eval_loss": 0.1254514902830124,
"eval_runtime": 3501.5361,
"eval_samples_per_second": 2.597,
"eval_steps_per_second": 0.081,
"eval_wer": 19.51153220566439,
"step": 5000
},
{
"epoch": 0.64,
"learning_rate": 6.281250000000001e-06,
"loss": 0.1604,
"step": 5025
},
{
"epoch": 0.65,
"learning_rate": 6.3125e-06,
"loss": 0.1633,
"step": 5050
},
{
"epoch": 0.65,
"learning_rate": 6.343750000000001e-06,
"loss": 0.1673,
"step": 5075
},
{
"epoch": 0.65,
"learning_rate": 6.375e-06,
"loss": 0.158,
"step": 5100
},
{
"epoch": 0.65,
"learning_rate": 6.406250000000001e-06,
"loss": 0.159,
"step": 5125
},
{
"epoch": 0.66,
"learning_rate": 6.437500000000001e-06,
"loss": 0.1651,
"step": 5150
},
{
"epoch": 0.66,
"learning_rate": 6.4687500000000005e-06,
"loss": 0.1645,
"step": 5175
},
{
"epoch": 0.66,
"learning_rate": 6.5000000000000004e-06,
"loss": 0.1592,
"step": 5200
},
{
"epoch": 0.67,
"learning_rate": 6.53125e-06,
"loss": 0.1549,
"step": 5225
},
{
"epoch": 0.67,
"learning_rate": 6.5625e-06,
"loss": 0.1646,
"step": 5250
},
{
"epoch": 0.67,
"learning_rate": 6.593750000000001e-06,
"loss": 0.1675,
"step": 5275
},
{
"epoch": 0.68,
"learning_rate": 6.625e-06,
"loss": 0.1605,
"step": 5300
},
{
"epoch": 0.68,
"learning_rate": 6.656250000000001e-06,
"loss": 0.1613,
"step": 5325
},
{
"epoch": 0.68,
"learning_rate": 6.6875e-06,
"loss": 0.1611,
"step": 5350
},
{
"epoch": 0.69,
"learning_rate": 6.718750000000001e-06,
"loss": 0.1555,
"step": 5375
},
{
"epoch": 0.69,
"learning_rate": 6.750000000000001e-06,
"loss": 0.1554,
"step": 5400
},
{
"epoch": 0.69,
"learning_rate": 6.7812500000000005e-06,
"loss": 0.1544,
"step": 5425
},
{
"epoch": 0.7,
"learning_rate": 6.8125e-06,
"loss": 0.1527,
"step": 5450
},
{
"epoch": 0.7,
"learning_rate": 6.84375e-06,
"loss": 0.1533,
"step": 5475
},
{
"epoch": 0.7,
"learning_rate": 6.875e-06,
"loss": 0.1593,
"step": 5500
},
{
"epoch": 0.71,
"learning_rate": 6.906250000000001e-06,
"loss": 0.1537,
"step": 5525
},
{
"epoch": 0.71,
"learning_rate": 6.9375e-06,
"loss": 0.1452,
"step": 5550
},
{
"epoch": 0.71,
"learning_rate": 6.968750000000001e-06,
"loss": 0.1592,
"step": 5575
},
{
"epoch": 0.72,
"learning_rate": 7e-06,
"loss": 0.1525,
"step": 5600
},
{
"epoch": 0.72,
"learning_rate": 7.031250000000001e-06,
"loss": 0.1603,
"step": 5625
},
{
"epoch": 0.72,
"learning_rate": 7.062500000000001e-06,
"loss": 0.152,
"step": 5650
},
{
"epoch": 0.73,
"learning_rate": 7.0937500000000005e-06,
"loss": 0.1547,
"step": 5675
},
{
"epoch": 0.73,
"learning_rate": 7.125e-06,
"loss": 0.1455,
"step": 5700
},
{
"epoch": 0.73,
"learning_rate": 7.15625e-06,
"loss": 0.1502,
"step": 5725
},
{
"epoch": 0.73,
"learning_rate": 7.1875e-06,
"loss": 0.1512,
"step": 5750
},
{
"epoch": 0.74,
"learning_rate": 7.218750000000001e-06,
"loss": 0.1522,
"step": 5775
},
{
"epoch": 0.74,
"learning_rate": 7.25e-06,
"loss": 0.1474,
"step": 5800
},
{
"epoch": 0.74,
"learning_rate": 7.281250000000001e-06,
"loss": 0.1557,
"step": 5825
},
{
"epoch": 0.75,
"learning_rate": 7.3125e-06,
"loss": 0.1534,
"step": 5850
},
{
"epoch": 0.75,
"learning_rate": 7.343750000000001e-06,
"loss": 0.1539,
"step": 5875
},
{
"epoch": 0.75,
"learning_rate": 7.375000000000001e-06,
"loss": 0.1448,
"step": 5900
},
{
"epoch": 0.76,
"learning_rate": 7.4062500000000005e-06,
"loss": 0.1499,
"step": 5925
},
{
"epoch": 0.76,
"learning_rate": 7.437500000000001e-06,
"loss": 0.1469,
"step": 5950
},
{
"epoch": 0.76,
"learning_rate": 7.46875e-06,
"loss": 0.1528,
"step": 5975
},
{
"epoch": 0.77,
"learning_rate": 7.500000000000001e-06,
"loss": 0.1504,
"step": 6000
},
{
"epoch": 0.77,
"eval_cer": 8.833290830940264,
"eval_loss": 0.11021216958761215,
"eval_runtime": 3540.1033,
"eval_samples_per_second": 2.569,
"eval_steps_per_second": 0.081,
"eval_wer": 17.191925969652313,
"step": 6000
},
{
"epoch": 0.77,
"learning_rate": 7.531250000000001e-06,
"loss": 0.1392,
"step": 6025
},
{
"epoch": 0.77,
"learning_rate": 7.5625e-06,
"loss": 0.1492,
"step": 6050
},
{
"epoch": 0.78,
"learning_rate": 7.593750000000001e-06,
"loss": 0.156,
"step": 6075
},
{
"epoch": 0.78,
"learning_rate": 7.625e-06,
"loss": 0.1434,
"step": 6100
},
{
"epoch": 0.78,
"learning_rate": 7.656250000000001e-06,
"loss": 0.1439,
"step": 6125
},
{
"epoch": 0.79,
"learning_rate": 7.6875e-06,
"loss": 0.1342,
"step": 6150
},
{
"epoch": 0.79,
"learning_rate": 7.71875e-06,
"loss": 0.142,
"step": 6175
},
{
"epoch": 0.79,
"learning_rate": 7.75e-06,
"loss": 0.1465,
"step": 6200
},
{
"epoch": 0.8,
"learning_rate": 7.78125e-06,
"loss": 0.1422,
"step": 6225
},
{
"epoch": 0.8,
"learning_rate": 7.8125e-06,
"loss": 0.1446,
"step": 6250
},
{
"epoch": 0.8,
"learning_rate": 7.843750000000001e-06,
"loss": 0.1387,
"step": 6275
},
{
"epoch": 0.81,
"learning_rate": 7.875e-06,
"loss": 0.1421,
"step": 6300
},
{
"epoch": 0.81,
"learning_rate": 7.90625e-06,
"loss": 0.1403,
"step": 6325
},
{
"epoch": 0.81,
"learning_rate": 7.9375e-06,
"loss": 0.1449,
"step": 6350
},
{
"epoch": 0.81,
"learning_rate": 7.96875e-06,
"loss": 0.1397,
"step": 6375
},
{
"epoch": 0.82,
"learning_rate": 8.000000000000001e-06,
"loss": 0.1364,
"step": 6400
},
{
"epoch": 0.82,
"learning_rate": 8.03125e-06,
"loss": 0.1402,
"step": 6425
},
{
"epoch": 0.82,
"learning_rate": 8.062500000000001e-06,
"loss": 0.1385,
"step": 6450
},
{
"epoch": 0.83,
"learning_rate": 8.09375e-06,
"loss": 0.1331,
"step": 6475
},
{
"epoch": 0.83,
"learning_rate": 8.125000000000001e-06,
"loss": 0.148,
"step": 6500
},
{
"epoch": 0.83,
"learning_rate": 8.156250000000002e-06,
"loss": 0.1371,
"step": 6525
},
{
"epoch": 0.84,
"learning_rate": 8.1875e-06,
"loss": 0.1382,
"step": 6550
},
{
"epoch": 0.84,
"learning_rate": 8.218750000000002e-06,
"loss": 0.1314,
"step": 6575
},
{
"epoch": 0.84,
"learning_rate": 8.25e-06,
"loss": 0.1377,
"step": 6600
},
{
"epoch": 0.85,
"learning_rate": 8.281250000000001e-06,
"loss": 0.1465,
"step": 6625
},
{
"epoch": 0.85,
"learning_rate": 8.3125e-06,
"loss": 0.1422,
"step": 6650
},
{
"epoch": 0.85,
"learning_rate": 8.343750000000001e-06,
"loss": 0.1376,
"step": 6675
},
{
"epoch": 0.86,
"learning_rate": 8.375e-06,
"loss": 0.1358,
"step": 6700
},
{
"epoch": 0.86,
"learning_rate": 8.406250000000001e-06,
"loss": 0.1366,
"step": 6725
},
{
"epoch": 0.86,
"learning_rate": 8.4375e-06,
"loss": 0.1392,
"step": 6750
},
{
"epoch": 0.87,
"learning_rate": 8.468750000000001e-06,
"loss": 0.1365,
"step": 6775
},
{
"epoch": 0.87,
"learning_rate": 8.5e-06,
"loss": 0.1349,
"step": 6800
},
{
"epoch": 0.87,
"learning_rate": 8.53125e-06,
"loss": 0.1342,
"step": 6825
},
{
"epoch": 0.88,
"learning_rate": 8.5625e-06,
"loss": 0.135,
"step": 6850
},
{
"epoch": 0.88,
"learning_rate": 8.59375e-06,
"loss": 0.1394,
"step": 6875
},
{
"epoch": 0.88,
"learning_rate": 8.625000000000001e-06,
"loss": 0.1336,
"step": 6900
},
{
"epoch": 0.88,
"learning_rate": 8.65625e-06,
"loss": 0.1319,
"step": 6925
},
{
"epoch": 0.89,
"learning_rate": 8.687500000000001e-06,
"loss": 0.1334,
"step": 6950
},
{
"epoch": 0.89,
"learning_rate": 8.71875e-06,
"loss": 0.1332,
"step": 6975
},
{
"epoch": 0.89,
"learning_rate": 8.750000000000001e-06,
"loss": 0.1259,
"step": 7000
},
{
"epoch": 0.89,
"eval_cer": 8.186330491216788,
"eval_loss": 0.10026197135448456,
"eval_runtime": 3538.7989,
"eval_samples_per_second": 2.57,
"eval_steps_per_second": 0.081,
"eval_wer": 15.857579064937932,
"step": 7000
},
{
"epoch": 0.9,
"learning_rate": 8.781250000000002e-06,
"loss": 0.1319,
"step": 7025
},
{
"epoch": 0.9,
"learning_rate": 8.8125e-06,
"loss": 0.1342,
"step": 7050
},
{
"epoch": 0.9,
"learning_rate": 8.843750000000002e-06,
"loss": 0.1339,
"step": 7075
},
{
"epoch": 0.91,
"learning_rate": 8.875e-06,
"loss": 0.1339,
"step": 7100
},
{
"epoch": 0.91,
"learning_rate": 8.906250000000001e-06,
"loss": 0.1378,
"step": 7125
},
{
"epoch": 0.91,
"learning_rate": 8.9375e-06,
"loss": 0.129,
"step": 7150
},
{
"epoch": 0.92,
"learning_rate": 8.968750000000001e-06,
"loss": 0.1325,
"step": 7175
},
{
"epoch": 0.92,
"learning_rate": 9e-06,
"loss": 0.1314,
"step": 7200
},
{
"epoch": 0.92,
"learning_rate": 9.031250000000001e-06,
"loss": 0.1332,
"step": 7225
},
{
"epoch": 0.93,
"learning_rate": 9.0625e-06,
"loss": 0.1344,
"step": 7250
},
{
"epoch": 0.93,
"learning_rate": 9.09375e-06,
"loss": 0.1306,
"step": 7275
},
{
"epoch": 0.93,
"learning_rate": 9.125e-06,
"loss": 0.1279,
"step": 7300
},
{
"epoch": 0.94,
"learning_rate": 9.15625e-06,
"loss": 0.1314,
"step": 7325
},
{
"epoch": 0.94,
"learning_rate": 9.1875e-06,
"loss": 0.1333,
"step": 7350
},
{
"epoch": 0.94,
"learning_rate": 9.21875e-06,
"loss": 0.1274,
"step": 7375
},
{
"epoch": 0.95,
"learning_rate": 9.250000000000001e-06,
"loss": 0.1291,
"step": 7400
},
{
"epoch": 0.95,
"learning_rate": 9.28125e-06,
"loss": 0.1297,
"step": 7425
},
{
"epoch": 0.95,
"learning_rate": 9.312500000000001e-06,
"loss": 0.1272,
"step": 7450
},
{
"epoch": 0.96,
"learning_rate": 9.34375e-06,
"loss": 0.1339,
"step": 7475
},
{
"epoch": 0.96,
"learning_rate": 9.375000000000001e-06,
"loss": 0.1258,
"step": 7500
},
{
"epoch": 0.96,
"learning_rate": 9.406250000000002e-06,
"loss": 0.1298,
"step": 7525
},
{
"epoch": 0.96,
"learning_rate": 9.4375e-06,
"loss": 0.1193,
"step": 7550
},
{
"epoch": 0.97,
"learning_rate": 9.468750000000001e-06,
"loss": 0.1283,
"step": 7575
},
{
"epoch": 0.97,
"learning_rate": 9.5e-06,
"loss": 0.128,
"step": 7600
},
{
"epoch": 0.97,
"learning_rate": 9.531250000000001e-06,
"loss": 0.1309,
"step": 7625
},
{
"epoch": 0.98,
"learning_rate": 9.562500000000002e-06,
"loss": 0.1305,
"step": 7650
},
{
"epoch": 0.98,
"learning_rate": 9.593750000000001e-06,
"loss": 0.1292,
"step": 7675
},
{
"epoch": 0.98,
"learning_rate": 9.625e-06,
"loss": 0.1267,
"step": 7700
},
{
"epoch": 0.99,
"learning_rate": 9.656250000000001e-06,
"loss": 0.128,
"step": 7725
},
{
"epoch": 0.99,
"learning_rate": 9.6875e-06,
"loss": 0.1246,
"step": 7750
},
{
"epoch": 0.99,
"learning_rate": 9.71875e-06,
"loss": 0.1237,
"step": 7775
},
{
"epoch": 1.0,
"learning_rate": 9.75e-06,
"loss": 0.1257,
"step": 7800
},
{
"epoch": 1.0,
"learning_rate": 9.78125e-06,
"loss": 0.1256,
"step": 7825
},
{
"epoch": 1.0,
"learning_rate": 9.8125e-06,
"loss": 0.1161,
"step": 7850
},
{
"epoch": 1.01,
"learning_rate": 9.84375e-06,
"loss": 0.1187,
"step": 7875
},
{
"epoch": 1.01,
"learning_rate": 9.875000000000001e-06,
"loss": 0.1149,
"step": 7900
},
{
"epoch": 1.01,
"learning_rate": 9.90625e-06,
"loss": 0.1188,
"step": 7925
},
{
"epoch": 1.02,
"learning_rate": 9.937500000000001e-06,
"loss": 0.1189,
"step": 7950
},
{
"epoch": 1.02,
"learning_rate": 9.96875e-06,
"loss": 0.1169,
"step": 7975
},
{
"epoch": 1.02,
"learning_rate": 1e-05,
"loss": 0.1184,
"step": 8000
},
{
"epoch": 1.02,
"eval_cer": 7.786807996363273,
"eval_loss": 0.09397807717323303,
"eval_runtime": 3511.0461,
"eval_samples_per_second": 2.59,
"eval_steps_per_second": 0.081,
"eval_wer": 14.910965681548896,
"step": 8000
},
{
"epoch": 1.03,
"learning_rate": 9.999984940186392e-06,
"loss": 0.1148,
"step": 8025
},
{
"epoch": 1.03,
"learning_rate": 9.999939760836287e-06,
"loss": 0.1164,
"step": 8050
},
{
"epoch": 1.03,
"learning_rate": 9.999864462221838e-06,
"loss": 0.116,
"step": 8075
},
{
"epoch": 1.04,
"learning_rate": 9.99975904479664e-06,
"loss": 0.1077,
"step": 8100
},
{
"epoch": 1.04,
"learning_rate": 9.999623509195724e-06,
"loss": 0.1159,
"step": 8125
},
{
"epoch": 1.04,
"learning_rate": 9.999457856235542e-06,
"loss": 0.1163,
"step": 8150
},
{
"epoch": 1.04,
"learning_rate": 9.999262086913975e-06,
"loss": 0.1169,
"step": 8175
},
{
"epoch": 1.05,
"learning_rate": 9.999036202410324e-06,
"loss": 0.1118,
"step": 8200
},
{
"epoch": 1.05,
"learning_rate": 9.998780204085303e-06,
"loss": 0.1194,
"step": 8225
},
{
"epoch": 1.05,
"learning_rate": 9.998494093481022e-06,
"loss": 0.1122,
"step": 8250
},
{
"epoch": 1.06,
"learning_rate": 9.998177872320993e-06,
"loss": 0.12,
"step": 8275
},
{
"epoch": 1.06,
"learning_rate": 9.997831542510107e-06,
"loss": 0.1114,
"step": 8300
},
{
"epoch": 1.06,
"learning_rate": 9.99745510613463e-06,
"loss": 0.1174,
"step": 8325
},
{
"epoch": 1.07,
"learning_rate": 9.997048565462188e-06,
"loss": 0.119,
"step": 8350
},
{
"epoch": 1.07,
"learning_rate": 9.996611922941748e-06,
"loss": 0.1112,
"step": 8375
},
{
"epoch": 1.07,
"learning_rate": 9.996145181203616e-06,
"loss": 0.1138,
"step": 8400
},
{
"epoch": 1.08,
"learning_rate": 9.995648343059407e-06,
"loss": 0.1145,
"step": 8425
},
{
"epoch": 1.08,
"learning_rate": 9.995121411502037e-06,
"loss": 0.1131,
"step": 8450
},
{
"epoch": 1.08,
"learning_rate": 9.994564389705702e-06,
"loss": 0.1117,
"step": 8475
},
{
"epoch": 1.09,
"learning_rate": 9.993977281025862e-06,
"loss": 0.1109,
"step": 8500
},
{
"epoch": 1.09,
"learning_rate": 9.993360088999216e-06,
"loss": 0.1162,
"step": 8525
},
{
"epoch": 1.09,
"learning_rate": 9.99271281734368e-06,
"loss": 0.1127,
"step": 8550
},
{
"epoch": 1.1,
"learning_rate": 9.992035469958371e-06,
"loss": 0.1097,
"step": 8575
},
{
"epoch": 1.1,
"learning_rate": 9.99132805092358e-06,
"loss": 0.1074,
"step": 8600
},
{
"epoch": 1.1,
"learning_rate": 9.990590564500745e-06,
"loss": 0.1146,
"step": 8625
},
{
"epoch": 1.11,
"learning_rate": 9.989823015132433e-06,
"loss": 0.1158,
"step": 8650
},
{
"epoch": 1.11,
"learning_rate": 9.989025407442301e-06,
"loss": 0.1124,
"step": 8675
},
{
"epoch": 1.11,
"learning_rate": 9.98819774623508e-06,
"loss": 0.1091,
"step": 8700
},
{
"epoch": 1.11,
"learning_rate": 9.987340036496535e-06,
"loss": 0.1108,
"step": 8725
},
{
"epoch": 1.12,
"learning_rate": 9.986452283393452e-06,
"loss": 0.1173,
"step": 8750
},
{
"epoch": 1.12,
"learning_rate": 9.985534492273585e-06,
"loss": 0.1094,
"step": 8775
},
{
"epoch": 1.12,
"learning_rate": 9.984586668665641e-06,
"loss": 0.1085,
"step": 8800
},
{
"epoch": 1.13,
"learning_rate": 9.983608818279237e-06,
"loss": 0.1097,
"step": 8825
},
{
"epoch": 1.13,
"learning_rate": 9.982600947004875e-06,
"loss": 0.1168,
"step": 8850
},
{
"epoch": 1.13,
"learning_rate": 9.98156306091389e-06,
"loss": 0.1158,
"step": 8875
},
{
"epoch": 1.14,
"learning_rate": 9.980495166258437e-06,
"loss": 0.1118,
"step": 8900
},
{
"epoch": 1.14,
"learning_rate": 9.97939726947143e-06,
"loss": 0.1107,
"step": 8925
},
{
"epoch": 1.14,
"learning_rate": 9.978269377166517e-06,
"loss": 0.1043,
"step": 8950
},
{
"epoch": 1.15,
"learning_rate": 9.97711149613804e-06,
"loss": 0.1084,
"step": 8975
},
{
"epoch": 1.15,
"learning_rate": 9.975923633360985e-06,
"loss": 0.1099,
"step": 9000
},
{
"epoch": 1.15,
"eval_cer": 7.367512649081763,
"eval_loss": 0.08846528828144073,
"eval_runtime": 3492.7097,
"eval_samples_per_second": 2.603,
"eval_steps_per_second": 0.082,
"eval_wer": 13.944392303086156,
"step": 9000
},
{
"epoch": 1.15,
"learning_rate": 9.974705795990952e-06,
"loss": 0.1055,
"step": 9025
},
{
"epoch": 1.16,
"learning_rate": 9.973457991364098e-06,
"loss": 0.1098,
"step": 9050
},
{
"epoch": 1.16,
"learning_rate": 9.97218022699711e-06,
"loss": 0.1034,
"step": 9075
},
{
"epoch": 1.16,
"learning_rate": 9.970872510587142e-06,
"loss": 0.1092,
"step": 9100
},
{
"epoch": 1.17,
"learning_rate": 9.969534850011782e-06,
"loss": 0.1123,
"step": 9125
},
{
"epoch": 1.17,
"learning_rate": 9.968167253328995e-06,
"loss": 0.1057,
"step": 9150
},
{
"epoch": 1.17,
"learning_rate": 9.966769728777084e-06,
"loss": 0.1057,
"step": 9175
},
{
"epoch": 1.18,
"learning_rate": 9.965342284774633e-06,
"loss": 0.1023,
"step": 9200
},
{
"epoch": 1.18,
"learning_rate": 9.963884929920455e-06,
"loss": 0.1139,
"step": 9225
},
{
"epoch": 1.18,
"learning_rate": 9.962397672993552e-06,
"loss": 0.1046,
"step": 9250
},
{
"epoch": 1.19,
"learning_rate": 9.960880522953043e-06,
"loss": 0.1124,
"step": 9275
},
{
"epoch": 1.19,
"learning_rate": 9.95933348893813e-06,
"loss": 0.1106,
"step": 9300
},
{
"epoch": 1.19,
"learning_rate": 9.957756580268033e-06,
"loss": 0.1044,
"step": 9325
},
{
"epoch": 1.19,
"learning_rate": 9.956149806441927e-06,
"loss": 0.1093,
"step": 9350
},
{
"epoch": 1.2,
"learning_rate": 9.9545131771389e-06,
"loss": 0.1138,
"step": 9375
},
{
"epoch": 1.2,
"learning_rate": 9.952846702217886e-06,
"loss": 0.0964,
"step": 9400
},
{
"epoch": 1.2,
"learning_rate": 9.951150391717605e-06,
"loss": 0.1111,
"step": 9425
},
{
"epoch": 1.21,
"learning_rate": 9.949424255856506e-06,
"loss": 0.108,
"step": 9450
},
{
"epoch": 1.21,
"learning_rate": 9.9476683050327e-06,
"loss": 0.1023,
"step": 9475
},
{
"epoch": 1.21,
"learning_rate": 9.945882549823906e-06,
"loss": 0.1008,
"step": 9500
},
{
"epoch": 1.22,
"learning_rate": 9.944067000987378e-06,
"loss": 0.1023,
"step": 9525
},
{
"epoch": 1.22,
"learning_rate": 9.94222166945985e-06,
"loss": 0.1067,
"step": 9550
},
{
"epoch": 1.22,
"learning_rate": 9.94034656635746e-06,
"loss": 0.1061,
"step": 9575
},
{
"epoch": 1.23,
"learning_rate": 9.938441702975689e-06,
"loss": 0.1062,
"step": 9600
},
{
"epoch": 1.23,
"learning_rate": 9.936507090789294e-06,
"loss": 0.108,
"step": 9625
},
{
"epoch": 1.23,
"learning_rate": 9.93454274145223e-06,
"loss": 0.1007,
"step": 9650
},
{
"epoch": 1.24,
"learning_rate": 9.932548666797595e-06,
"loss": 0.1091,
"step": 9675
},
{
"epoch": 1.24,
"learning_rate": 9.930524878837544e-06,
"loss": 0.1013,
"step": 9700
},
{
"epoch": 1.24,
"learning_rate": 9.928471389763226e-06,
"loss": 0.101,
"step": 9725
},
{
"epoch": 1.25,
"learning_rate": 9.926388211944707e-06,
"loss": 0.1045,
"step": 9750
},
{
"epoch": 1.25,
"learning_rate": 9.924275357930893e-06,
"loss": 0.1036,
"step": 9775
},
{
"epoch": 1.25,
"learning_rate": 9.922132840449459e-06,
"loss": 0.107,
"step": 9800
},
{
"epoch": 1.26,
"learning_rate": 9.919960672406772e-06,
"loss": 0.1017,
"step": 9825
},
{
"epoch": 1.26,
"learning_rate": 9.917758866887808e-06,
"loss": 0.1042,
"step": 9850
},
{
"epoch": 1.26,
"learning_rate": 9.915527437156083e-06,
"loss": 0.1045,
"step": 9875
},
{
"epoch": 1.27,
"learning_rate": 9.91326639665356e-06,
"loss": 0.0966,
"step": 9900
},
{
"epoch": 1.27,
"learning_rate": 9.910975759000578e-06,
"loss": 0.1,
"step": 9925
},
{
"epoch": 1.27,
"learning_rate": 9.908655537995772e-06,
"loss": 0.1025,
"step": 9950
},
{
"epoch": 1.27,
"learning_rate": 9.906305747615976e-06,
"loss": 0.1046,
"step": 9975
},
{
"epoch": 1.28,
"learning_rate": 9.903926402016153e-06,
"loss": 0.1075,
"step": 10000
},
{
"epoch": 1.28,
"eval_cer": 6.964848673001371,
"eval_loss": 0.08299314230680466,
"eval_runtime": 3511.9689,
"eval_samples_per_second": 2.589,
"eval_steps_per_second": 0.081,
"eval_wer": 13.200776316404141,
"step": 10000
},
{
"epoch": 1.28,
"learning_rate": 9.901517515529303e-06,
"loss": 0.1068,
"step": 10025
},
{
"epoch": 1.28,
"learning_rate": 9.899079102666382e-06,
"loss": 0.1029,
"step": 10050
},
{
"epoch": 1.29,
"learning_rate": 9.896611178116205e-06,
"loss": 0.1051,
"step": 10075
},
{
"epoch": 1.29,
"learning_rate": 9.894113756745362e-06,
"loss": 0.1013,
"step": 10100
},
{
"epoch": 1.29,
"learning_rate": 9.891586853598139e-06,
"loss": 0.0986,
"step": 10125
},
{
"epoch": 1.3,
"learning_rate": 9.88903048389641e-06,
"loss": 0.099,
"step": 10150
},
{
"epoch": 1.3,
"learning_rate": 9.886444663039552e-06,
"loss": 0.0964,
"step": 10175
},
{
"epoch": 1.3,
"learning_rate": 9.883829406604363e-06,
"loss": 0.1044,
"step": 10200
},
{
"epoch": 1.31,
"learning_rate": 9.881184730344949e-06,
"loss": 0.1064,
"step": 10225
},
{
"epoch": 1.31,
"learning_rate": 9.878510650192644e-06,
"loss": 0.0994,
"step": 10250
},
{
"epoch": 1.31,
"learning_rate": 9.875807182255905e-06,
"loss": 0.0986,
"step": 10275
},
{
"epoch": 1.32,
"learning_rate": 9.873074342820225e-06,
"loss": 0.1073,
"step": 10300
},
{
"epoch": 1.32,
"learning_rate": 9.870312148348024e-06,
"loss": 0.1011,
"step": 10325
},
{
"epoch": 1.32,
"learning_rate": 9.867520615478554e-06,
"loss": 0.0983,
"step": 10350
},
{
"epoch": 1.33,
"learning_rate": 9.864699761027801e-06,
"loss": 0.1016,
"step": 10375
},
{
"epoch": 1.33,
"learning_rate": 9.861849601988384e-06,
"loss": 0.1039,
"step": 10400
},
{
"epoch": 1.33,
"learning_rate": 9.858970155529447e-06,
"loss": 0.0979,
"step": 10425
},
{
"epoch": 1.34,
"learning_rate": 9.85606143899656e-06,
"loss": 0.103,
"step": 10450
},
{
"epoch": 1.34,
"learning_rate": 9.853123469911616e-06,
"loss": 0.1006,
"step": 10475
},
{
"epoch": 1.34,
"learning_rate": 9.850156265972722e-06,
"loss": 0.1008,
"step": 10500
},
{
"epoch": 1.35,
"learning_rate": 9.847159845054091e-06,
"loss": 0.0984,
"step": 10525
},
{
"epoch": 1.35,
"learning_rate": 9.844134225205941e-06,
"loss": 0.1007,
"step": 10550
},
{
"epoch": 1.35,
"learning_rate": 9.84107942465438e-06,
"loss": 0.0988,
"step": 10575
},
{
"epoch": 1.35,
"learning_rate": 9.8379954618013e-06,
"loss": 0.0972,
"step": 10600
},
{
"epoch": 1.36,
"learning_rate": 9.834882355224261e-06,
"loss": 0.0946,
"step": 10625
},
{
"epoch": 1.36,
"learning_rate": 9.831740123676387e-06,
"loss": 0.1059,
"step": 10650
},
{
"epoch": 1.36,
"learning_rate": 9.828568786086245e-06,
"loss": 0.1015,
"step": 10675
},
{
"epoch": 1.37,
"learning_rate": 9.825368361557738e-06,
"loss": 0.102,
"step": 10700
},
{
"epoch": 1.37,
"learning_rate": 9.822138869369983e-06,
"loss": 0.0953,
"step": 10725
},
{
"epoch": 1.37,
"learning_rate": 9.8188803289772e-06,
"loss": 0.0993,
"step": 10750
},
{
"epoch": 1.38,
"learning_rate": 9.815592760008594e-06,
"loss": 0.0949,
"step": 10775
},
{
"epoch": 1.38,
"learning_rate": 9.812276182268236e-06,
"loss": 0.1005,
"step": 10800
},
{
"epoch": 1.38,
"learning_rate": 9.808930615734943e-06,
"loss": 0.0954,
"step": 10825
},
{
"epoch": 1.39,
"learning_rate": 9.805556080562159e-06,
"loss": 0.1012,
"step": 10850
},
{
"epoch": 1.39,
"learning_rate": 9.80215259707783e-06,
"loss": 0.0956,
"step": 10875
},
{
"epoch": 1.39,
"learning_rate": 9.798720185784288e-06,
"loss": 0.0961,
"step": 10900
},
{
"epoch": 1.4,
"learning_rate": 9.795258867358122e-06,
"loss": 0.1006,
"step": 10925
},
{
"epoch": 1.4,
"learning_rate": 9.791768662650059e-06,
"loss": 0.0962,
"step": 10950
},
{
"epoch": 1.4,
"learning_rate": 9.788249592684826e-06,
"loss": 0.1013,
"step": 10975
},
{
"epoch": 1.41,
"learning_rate": 9.784701678661045e-06,
"loss": 0.095,
"step": 11000
},
{
"epoch": 1.41,
"eval_cer": 6.696898803650031,
"eval_loss": 0.07891383767127991,
"eval_runtime": 3505.5381,
"eval_samples_per_second": 2.594,
"eval_steps_per_second": 0.081,
"eval_wer": 12.677569637025366,
"step": 11000
},
{
"epoch": 1.41,
"learning_rate": 9.78112494195108e-06,
"loss": 0.0961,
"step": 11025
},
{
"epoch": 1.41,
"learning_rate": 9.777519404100933e-06,
"loss": 0.098,
"step": 11050
},
{
"epoch": 1.42,
"learning_rate": 9.773885086830086e-06,
"loss": 0.0962,
"step": 11075
},
{
"epoch": 1.42,
"learning_rate": 9.770222012031404e-06,
"loss": 0.0912,
"step": 11100
},
{
"epoch": 1.42,
"learning_rate": 9.766530201770969e-06,
"loss": 0.1048,
"step": 11125
},
{
"epoch": 1.42,
"learning_rate": 9.762809678287977e-06,
"loss": 0.0918,
"step": 11150
},
{
"epoch": 1.43,
"learning_rate": 9.75906046399458e-06,
"loss": 0.1024,
"step": 11175
},
{
"epoch": 1.43,
"learning_rate": 9.755282581475769e-06,
"loss": 0.0917,
"step": 11200
},
{
"epoch": 1.43,
"learning_rate": 9.751476053489223e-06,
"loss": 0.0974,
"step": 11225
},
{
"epoch": 1.44,
"learning_rate": 9.747640902965185e-06,
"loss": 0.0944,
"step": 11250
},
{
"epoch": 1.44,
"learning_rate": 9.743777153006312e-06,
"loss": 0.0983,
"step": 11275
},
{
"epoch": 1.44,
"learning_rate": 9.739884826887554e-06,
"loss": 0.0912,
"step": 11300
},
{
"epoch": 1.45,
"learning_rate": 9.735963948055984e-06,
"loss": 0.097,
"step": 11325
},
{
"epoch": 1.45,
"learning_rate": 9.73201454013069e-06,
"loss": 0.0972,
"step": 11350
},
{
"epoch": 1.45,
"learning_rate": 9.728036626902607e-06,
"loss": 0.0927,
"step": 11375
},
{
"epoch": 1.46,
"learning_rate": 9.72403023233439e-06,
"loss": 0.0953,
"step": 11400
},
{
"epoch": 1.46,
"learning_rate": 9.719995380560261e-06,
"loss": 0.0992,
"step": 11425
},
{
"epoch": 1.46,
"learning_rate": 9.715932095885867e-06,
"loss": 0.0958,
"step": 11450
},
{
"epoch": 1.47,
"learning_rate": 9.71184040278813e-06,
"loss": 0.0922,
"step": 11475
},
{
"epoch": 1.47,
"learning_rate": 9.707720325915105e-06,
"loss": 0.0926,
"step": 11500
},
{
"epoch": 1.47,
"learning_rate": 9.703571890085828e-06,
"loss": 0.0938,
"step": 11525
},
{
"epoch": 1.48,
"learning_rate": 9.699395120290166e-06,
"loss": 0.0956,
"step": 11550
},
{
"epoch": 1.48,
"learning_rate": 9.69519004168867e-06,
"loss": 0.0899,
"step": 11575
},
{
"epoch": 1.48,
"learning_rate": 9.690956679612422e-06,
"loss": 0.1021,
"step": 11600
},
{
"epoch": 1.49,
"learning_rate": 9.686695059562875e-06,
"loss": 0.0921,
"step": 11625
},
{
"epoch": 1.49,
"learning_rate": 9.682405207211714e-06,
"loss": 0.0901,
"step": 11650
},
{
"epoch": 1.49,
"learning_rate": 9.678087148400688e-06,
"loss": 0.0947,
"step": 11675
},
{
"epoch": 1.5,
"learning_rate": 9.673740909141463e-06,
"loss": 0.0934,
"step": 11700
},
{
"epoch": 1.5,
"learning_rate": 9.669366515615458e-06,
"loss": 0.0929,
"step": 11725
},
{
"epoch": 1.5,
"learning_rate": 9.664963994173695e-06,
"loss": 0.0933,
"step": 11750
},
{
"epoch": 1.5,
"learning_rate": 9.660533371336636e-06,
"loss": 0.0907,
"step": 11775
},
{
"epoch": 1.51,
"learning_rate": 9.656074673794018e-06,
"loss": 0.0904,
"step": 11800
},
{
"epoch": 1.51,
"learning_rate": 9.65158792840471e-06,
"loss": 0.0901,
"step": 11825
},
{
"epoch": 1.51,
"learning_rate": 9.647073162196524e-06,
"loss": 0.1007,
"step": 11850
},
{
"epoch": 1.52,
"learning_rate": 9.64253040236608e-06,
"loss": 0.0898,
"step": 11875
},
{
"epoch": 1.52,
"learning_rate": 9.637959676278621e-06,
"loss": 0.0962,
"step": 11900
},
{
"epoch": 1.52,
"learning_rate": 9.633361011467864e-06,
"loss": 0.0917,
"step": 11925
},
{
"epoch": 1.53,
"learning_rate": 9.62873443563582e-06,
"loss": 0.0919,
"step": 11950
},
{
"epoch": 1.53,
"learning_rate": 9.62407997665264e-06,
"loss": 0.0911,
"step": 11975
},
{
"epoch": 1.53,
"learning_rate": 9.619397662556434e-06,
"loss": 0.0943,
"step": 12000
},
{
"epoch": 1.53,
"eval_cer": 6.376467718508498,
"eval_loss": 0.07664290815591812,
"eval_runtime": 3518.3769,
"eval_samples_per_second": 2.584,
"eval_steps_per_second": 0.081,
"eval_wer": 11.989586828101974,
"step": 12000
},
{
"epoch": 1.54,
"learning_rate": 9.614687521553116e-06,
"loss": 0.0986,
"step": 12025
},
{
"epoch": 1.54,
"learning_rate": 9.609949582016223e-06,
"loss": 0.0899,
"step": 12050
},
{
"epoch": 1.54,
"learning_rate": 9.605183872486749e-06,
"loss": 0.0919,
"step": 12075
},
{
"epoch": 1.55,
"learning_rate": 9.600390421672976e-06,
"loss": 0.0925,
"step": 12100
},
{
"epoch": 1.55,
"learning_rate": 9.59556925845029e-06,
"loss": 0.089,
"step": 12125
},
{
"epoch": 1.55,
"learning_rate": 9.590720411861022e-06,
"loss": 0.0899,
"step": 12150
},
{
"epoch": 1.56,
"learning_rate": 9.58584391111426e-06,
"loss": 0.085,
"step": 12175
},
{
"epoch": 1.56,
"learning_rate": 9.58093978558568e-06,
"loss": 0.0923,
"step": 12200
},
{
"epoch": 1.56,
"learning_rate": 9.576008064817373e-06,
"loss": 0.0892,
"step": 12225
},
{
"epoch": 1.57,
"learning_rate": 9.571048778517655e-06,
"loss": 0.0851,
"step": 12250
},
{
"epoch": 1.57,
"learning_rate": 9.566061956560895e-06,
"loss": 0.0857,
"step": 12275
},
{
"epoch": 1.57,
"learning_rate": 9.561047628987338e-06,
"loss": 0.0954,
"step": 12300
},
{
"epoch": 1.58,
"learning_rate": 9.556005826002922e-06,
"loss": 0.0934,
"step": 12325
},
{
"epoch": 1.58,
"learning_rate": 9.55093657797909e-06,
"loss": 0.0868,
"step": 12350
},
{
"epoch": 1.58,
"learning_rate": 9.545839915452612e-06,
"loss": 0.0913,
"step": 12375
},
{
"epoch": 1.58,
"learning_rate": 9.540715869125407e-06,
"loss": 0.0942,
"step": 12400
},
{
"epoch": 1.59,
"learning_rate": 9.535564469864347e-06,
"loss": 0.0885,
"step": 12425
},
{
"epoch": 1.59,
"learning_rate": 9.530385748701074e-06,
"loss": 0.0926,
"step": 12450
},
{
"epoch": 1.59,
"learning_rate": 9.525179736831822e-06,
"loss": 0.0929,
"step": 12475
},
{
"epoch": 1.6,
"learning_rate": 9.519946465617217e-06,
"loss": 0.0875,
"step": 12500
},
{
"epoch": 1.6,
"learning_rate": 9.514685966582096e-06,
"loss": 0.0967,
"step": 12525
},
{
"epoch": 1.6,
"learning_rate": 9.509398271415308e-06,
"loss": 0.0907,
"step": 12550
},
{
"epoch": 1.61,
"learning_rate": 9.50408341196954e-06,
"loss": 0.0926,
"step": 12575
},
{
"epoch": 1.61,
"learning_rate": 9.498741420261109e-06,
"loss": 0.0864,
"step": 12600
},
{
"epoch": 1.61,
"learning_rate": 9.49337232846977e-06,
"loss": 0.0905,
"step": 12625
},
{
"epoch": 1.62,
"learning_rate": 9.487976168938535e-06,
"loss": 0.0913,
"step": 12650
},
{
"epoch": 1.62,
"learning_rate": 9.482552974173467e-06,
"loss": 0.0873,
"step": 12675
},
{
"epoch": 1.62,
"learning_rate": 9.477102776843486e-06,
"loss": 0.0873,
"step": 12700
},
{
"epoch": 1.63,
"learning_rate": 9.471625609780173e-06,
"loss": 0.0941,
"step": 12725
},
{
"epoch": 1.63,
"learning_rate": 9.466121505977577e-06,
"loss": 0.0951,
"step": 12750
},
{
"epoch": 1.63,
"learning_rate": 9.460590498592007e-06,
"loss": 0.0911,
"step": 12775
},
{
"epoch": 1.64,
"learning_rate": 9.45503262094184e-06,
"loss": 0.0866,
"step": 12800
},
{
"epoch": 1.64,
"learning_rate": 9.449447906507316e-06,
"loss": 0.0905,
"step": 12825
},
{
"epoch": 1.64,
"learning_rate": 9.443836388930339e-06,
"loss": 0.0869,
"step": 12850
},
{
"epoch": 1.65,
"learning_rate": 9.438198102014271e-06,
"loss": 0.0909,
"step": 12875
},
{
"epoch": 1.65,
"learning_rate": 9.432533079723734e-06,
"loss": 0.0937,
"step": 12900
},
{
"epoch": 1.65,
"learning_rate": 9.426841356184398e-06,
"loss": 0.0857,
"step": 12925
},
{
"epoch": 1.65,
"learning_rate": 9.421122965682782e-06,
"loss": 0.0909,
"step": 12950
},
{
"epoch": 1.66,
"learning_rate": 9.415377942666044e-06,
"loss": 0.0902,
"step": 12975
},
{
"epoch": 1.66,
"learning_rate": 9.409606321741776e-06,
"loss": 0.0923,
"step": 13000
},
{
"epoch": 1.66,
"eval_cer": 6.178369608201853,
"eval_loss": 0.07307101041078568,
"eval_runtime": 3502.2615,
"eval_samples_per_second": 2.596,
"eval_steps_per_second": 0.081,
"eval_wer": 11.720339234980104,
"step": 13000
},
{
"epoch": 1.66,
"learning_rate": 9.40380813767779e-06,
"loss": 0.0894,
"step": 13025
},
{
"epoch": 1.67,
"learning_rate": 9.397983425401915e-06,
"loss": 0.0831,
"step": 13050
},
{
"epoch": 1.67,
"learning_rate": 9.392132220001784e-06,
"loss": 0.089,
"step": 13075
},
{
"epoch": 1.67,
"learning_rate": 9.386254556724622e-06,
"loss": 0.0835,
"step": 13100
},
{
"epoch": 1.68,
"learning_rate": 9.380350470977033e-06,
"loss": 0.0881,
"step": 13125
},
{
"epoch": 1.68,
"learning_rate": 9.374419998324792e-06,
"loss": 0.089,
"step": 13150
},
{
"epoch": 1.68,
"learning_rate": 9.368463174492623e-06,
"loss": 0.0897,
"step": 13175
},
{
"epoch": 1.69,
"learning_rate": 9.362480035363987e-06,
"loss": 0.0859,
"step": 13200
},
{
"epoch": 1.69,
"learning_rate": 9.356470616980868e-06,
"loss": 0.0952,
"step": 13225
},
{
"epoch": 1.69,
"learning_rate": 9.350434955543557e-06,
"loss": 0.0873,
"step": 13250
},
{
"epoch": 1.7,
"learning_rate": 9.344373087410428e-06,
"loss": 0.0855,
"step": 13275
},
{
"epoch": 1.7,
"learning_rate": 9.338285049097722e-06,
"loss": 0.0859,
"step": 13300
},
{
"epoch": 1.7,
"learning_rate": 9.332170877279325e-06,
"loss": 0.0904,
"step": 13325
},
{
"epoch": 1.71,
"learning_rate": 9.326030608786558e-06,
"loss": 0.0838,
"step": 13350
},
{
"epoch": 1.71,
"learning_rate": 9.319864280607935e-06,
"loss": 0.0885,
"step": 13375
},
{
"epoch": 1.71,
"learning_rate": 9.31367192988896e-06,
"loss": 0.0876,
"step": 13400
},
{
"epoch": 1.72,
"learning_rate": 9.307453593931893e-06,
"loss": 0.0902,
"step": 13425
},
{
"epoch": 1.72,
"learning_rate": 9.301209310195523e-06,
"loss": 0.0822,
"step": 13450
},
{
"epoch": 1.72,
"learning_rate": 9.294939116294952e-06,
"loss": 0.0881,
"step": 13475
},
{
"epoch": 1.73,
"learning_rate": 9.288643050001362e-06,
"loss": 0.0871,
"step": 13500
},
{
"epoch": 1.73,
"learning_rate": 9.282321149241783e-06,
"loss": 0.0895,
"step": 13525
},
{
"epoch": 1.73,
"learning_rate": 9.275973452098877e-06,
"loss": 0.0879,
"step": 13550
},
{
"epoch": 1.73,
"learning_rate": 9.269599996810696e-06,
"loss": 0.0874,
"step": 13575
},
{
"epoch": 1.74,
"learning_rate": 9.263200821770462e-06,
"loss": 0.0849,
"step": 13600
},
{
"epoch": 1.74,
"learning_rate": 9.256775965526327e-06,
"loss": 0.0847,
"step": 13625
},
{
"epoch": 1.74,
"learning_rate": 9.250325466781145e-06,
"loss": 0.0808,
"step": 13650
},
{
"epoch": 1.75,
"learning_rate": 9.243849364392241e-06,
"loss": 0.0892,
"step": 13675
},
{
"epoch": 1.75,
"learning_rate": 9.237347697371173e-06,
"loss": 0.0894,
"step": 13700
},
{
"epoch": 1.75,
"learning_rate": 9.230820504883498e-06,
"loss": 0.0819,
"step": 13725
},
{
"epoch": 1.76,
"learning_rate": 9.224267826248536e-06,
"loss": 0.0858,
"step": 13750
},
{
"epoch": 1.76,
"learning_rate": 9.217689700939137e-06,
"loss": 0.0842,
"step": 13775
},
{
"epoch": 1.76,
"learning_rate": 9.211086168581433e-06,
"loss": 0.0834,
"step": 13800
},
{
"epoch": 1.77,
"learning_rate": 9.204457268954615e-06,
"loss": 0.0878,
"step": 13825
},
{
"epoch": 1.77,
"learning_rate": 9.19780304199068e-06,
"loss": 0.0828,
"step": 13850
},
{
"epoch": 1.77,
"learning_rate": 9.19112352777419e-06,
"loss": 0.083,
"step": 13875
},
{
"epoch": 1.78,
"learning_rate": 9.184418766542046e-06,
"loss": 0.0841,
"step": 13900
},
{
"epoch": 1.78,
"learning_rate": 9.17768879868323e-06,
"loss": 0.0829,
"step": 13925
},
{
"epoch": 1.78,
"learning_rate": 9.170933664738563e-06,
"loss": 0.085,
"step": 13950
},
{
"epoch": 1.79,
"learning_rate": 9.16415340540047e-06,
"loss": 0.084,
"step": 13975
},
{
"epoch": 1.79,
"learning_rate": 9.157348061512728e-06,
"loss": 0.0824,
"step": 14000
},
{
"epoch": 1.79,
"eval_cer": 5.926681524024939,
"eval_loss": 0.06988836824893951,
"eval_runtime": 3512.2323,
"eval_samples_per_second": 2.589,
"eval_steps_per_second": 0.081,
"eval_wer": 11.163158095901405,
"step": 14000
},
{
"epoch": 1.79,
"learning_rate": 9.15051767407022e-06,
"loss": 0.0858,
"step": 14025
},
{
"epoch": 1.8,
"learning_rate": 9.143662284218691e-06,
"loss": 0.0839,
"step": 14050
},
{
"epoch": 1.8,
"learning_rate": 9.1367819332545e-06,
"loss": 0.0849,
"step": 14075
},
{
"epoch": 1.8,
"learning_rate": 9.129876662624366e-06,
"loss": 0.0789,
"step": 14100
},
{
"epoch": 1.81,
"learning_rate": 9.122946513925128e-06,
"loss": 0.0847,
"step": 14125
},
{
"epoch": 1.81,
"learning_rate": 9.11599152890348e-06,
"loss": 0.0849,
"step": 14150
},
{
"epoch": 1.81,
"learning_rate": 9.109011749455738e-06,
"loss": 0.0848,
"step": 14175
},
{
"epoch": 1.81,
"learning_rate": 9.102007217627568e-06,
"loss": 0.085,
"step": 14200
},
{
"epoch": 1.82,
"learning_rate": 9.094977975613752e-06,
"loss": 0.0858,
"step": 14225
},
{
"epoch": 1.82,
"learning_rate": 9.08792406575792e-06,
"loss": 0.0888,
"step": 14250
},
{
"epoch": 1.82,
"learning_rate": 9.080845530552294e-06,
"loss": 0.0803,
"step": 14275
},
{
"epoch": 1.83,
"learning_rate": 9.073742412637448e-06,
"loss": 0.0849,
"step": 14300
},
{
"epoch": 1.83,
"learning_rate": 9.06661475480203e-06,
"loss": 0.0782,
"step": 14325
},
{
"epoch": 1.83,
"learning_rate": 9.059462599982525e-06,
"loss": 0.0815,
"step": 14350
},
{
"epoch": 1.84,
"learning_rate": 9.052285991262975e-06,
"loss": 0.0841,
"step": 14375
},
{
"epoch": 1.84,
"learning_rate": 9.045084971874738e-06,
"loss": 0.0787,
"step": 14400
},
{
"epoch": 1.84,
"learning_rate": 9.037859585196218e-06,
"loss": 0.0794,
"step": 14425
},
{
"epoch": 1.85,
"learning_rate": 9.030609874752604e-06,
"loss": 0.0828,
"step": 14450
},
{
"epoch": 1.85,
"learning_rate": 9.023335884215616e-06,
"loss": 0.0878,
"step": 14475
},
{
"epoch": 1.85,
"learning_rate": 9.016037657403225e-06,
"loss": 0.0784,
"step": 14500
},
{
"epoch": 1.86,
"learning_rate": 9.00871523827941e-06,
"loss": 0.0857,
"step": 14525
},
{
"epoch": 1.86,
"learning_rate": 9.001368670953872e-06,
"loss": 0.0816,
"step": 14550
},
{
"epoch": 1.86,
"learning_rate": 8.99399799968179e-06,
"loss": 0.0837,
"step": 14575
},
{
"epoch": 1.87,
"learning_rate": 8.986603268863536e-06,
"loss": 0.0856,
"step": 14600
},
{
"epoch": 1.87,
"learning_rate": 8.979184523044419e-06,
"loss": 0.0815,
"step": 14625
},
{
"epoch": 1.87,
"learning_rate": 8.971741806914409e-06,
"loss": 0.0835,
"step": 14650
},
{
"epoch": 1.88,
"learning_rate": 8.964275165307871e-06,
"loss": 0.0811,
"step": 14675
},
{
"epoch": 1.88,
"learning_rate": 8.956784643203303e-06,
"loss": 0.0847,
"step": 14700
},
{
"epoch": 1.88,
"learning_rate": 8.949270285723047e-06,
"loss": 0.0819,
"step": 14725
},
{
"epoch": 1.88,
"learning_rate": 8.941732138133032e-06,
"loss": 0.0835,
"step": 14750
},
{
"epoch": 1.89,
"learning_rate": 8.9341702458425e-06,
"loss": 0.0822,
"step": 14775
},
{
"epoch": 1.89,
"learning_rate": 8.926584654403725e-06,
"loss": 0.08,
"step": 14800
},
{
"epoch": 1.89,
"learning_rate": 8.918975409511745e-06,
"loss": 0.0784,
"step": 14825
},
{
"epoch": 1.9,
"learning_rate": 8.911342557004084e-06,
"loss": 0.0867,
"step": 14850
},
{
"epoch": 1.9,
"learning_rate": 8.903686142860473e-06,
"loss": 0.0823,
"step": 14875
},
{
"epoch": 1.9,
"learning_rate": 8.896006213202584e-06,
"loss": 0.0778,
"step": 14900
},
{
"epoch": 1.91,
"learning_rate": 8.88830281429374e-06,
"loss": 0.0896,
"step": 14925
},
{
"epoch": 1.91,
"learning_rate": 8.88057599253864e-06,
"loss": 0.0767,
"step": 14950
},
{
"epoch": 1.91,
"learning_rate": 8.872825794483083e-06,
"loss": 0.0796,
"step": 14975
},
{
"epoch": 1.92,
"learning_rate": 8.865052266813686e-06,
"loss": 0.0756,
"step": 15000
},
{
"epoch": 1.92,
"eval_cer": 5.6304583236317,
"eval_loss": 0.06829311698675156,
"eval_runtime": 3516.4512,
"eval_samples_per_second": 2.586,
"eval_steps_per_second": 0.081,
"eval_wer": 10.632731843836396,
"step": 15000
},
{
"epoch": 1.92,
"learning_rate": 8.857255456357597e-06,
"loss": 0.0774,
"step": 15025
},
{
"epoch": 1.92,
"learning_rate": 8.849435410082224e-06,
"loss": 0.0822,
"step": 15050
},
{
"epoch": 1.93,
"learning_rate": 8.841592175094939e-06,
"loss": 0.0803,
"step": 15075
},
{
"epoch": 1.93,
"learning_rate": 8.833725798642809e-06,
"loss": 0.0873,
"step": 15100
},
{
"epoch": 1.93,
"learning_rate": 8.825836328112296e-06,
"loss": 0.0787,
"step": 15125
},
{
"epoch": 1.94,
"learning_rate": 8.817923811028984e-06,
"loss": 0.0762,
"step": 15150
},
{
"epoch": 1.94,
"learning_rate": 8.809988295057284e-06,
"loss": 0.0805,
"step": 15175
},
{
"epoch": 1.94,
"learning_rate": 8.802029828000157e-06,
"loss": 0.0742,
"step": 15200
},
{
"epoch": 1.95,
"learning_rate": 8.79404845779881e-06,
"loss": 0.0814,
"step": 15225
},
{
"epoch": 1.95,
"learning_rate": 8.786044232532423e-06,
"loss": 0.084,
"step": 15250
},
{
"epoch": 1.95,
"learning_rate": 8.778017200417856e-06,
"loss": 0.0787,
"step": 15275
},
{
"epoch": 1.96,
"learning_rate": 8.769967409809348e-06,
"loss": 0.0739,
"step": 15300
},
{
"epoch": 1.96,
"learning_rate": 8.76189490919824e-06,
"loss": 0.081,
"step": 15325
},
{
"epoch": 1.96,
"learning_rate": 8.753799747212672e-06,
"loss": 0.0782,
"step": 15350
},
{
"epoch": 1.96,
"learning_rate": 8.745681972617298e-06,
"loss": 0.0873,
"step": 15375
},
{
"epoch": 1.97,
"learning_rate": 8.737541634312985e-06,
"loss": 0.079,
"step": 15400
},
{
"epoch": 1.97,
"learning_rate": 8.729378781336524e-06,
"loss": 0.0782,
"step": 15425
},
{
"epoch": 1.97,
"learning_rate": 8.721193462860335e-06,
"loss": 0.0785,
"step": 15450
},
{
"epoch": 1.98,
"learning_rate": 8.712985728192167e-06,
"loss": 0.0827,
"step": 15475
},
{
"epoch": 1.98,
"learning_rate": 8.704755626774796e-06,
"loss": 0.0763,
"step": 15500
},
{
"epoch": 1.98,
"learning_rate": 8.696503208185745e-06,
"loss": 0.0782,
"step": 15525
},
{
"epoch": 1.99,
"learning_rate": 8.688228522136966e-06,
"loss": 0.0783,
"step": 15550
},
{
"epoch": 1.99,
"learning_rate": 8.67993161847455e-06,
"loss": 0.0739,
"step": 15575
},
{
"epoch": 1.99,
"learning_rate": 8.671612547178428e-06,
"loss": 0.0824,
"step": 15600
},
{
"epoch": 2.0,
"learning_rate": 8.663271358362064e-06,
"loss": 0.0828,
"step": 15625
},
{
"epoch": 2.0,
"learning_rate": 8.65490810227216e-06,
"loss": 0.0775,
"step": 15650
},
{
"epoch": 2.0,
"learning_rate": 8.646522829288343e-06,
"loss": 0.0694,
"step": 15675
},
{
"epoch": 2.01,
"learning_rate": 8.638115589922875e-06,
"loss": 0.0685,
"step": 15700
},
{
"epoch": 2.01,
"learning_rate": 8.62968643482034e-06,
"loss": 0.0637,
"step": 15725
},
{
"epoch": 2.01,
"learning_rate": 8.621235414757337e-06,
"loss": 0.0626,
"step": 15750
},
{
"epoch": 2.02,
"learning_rate": 8.61276258064218e-06,
"loss": 0.0656,
"step": 15775
},
{
"epoch": 2.02,
"learning_rate": 8.604267983514595e-06,
"loss": 0.0634,
"step": 15800
},
{
"epoch": 2.02,
"learning_rate": 8.595751674545398e-06,
"loss": 0.0716,
"step": 15825
},
{
"epoch": 2.03,
"learning_rate": 8.587213705036202e-06,
"loss": 0.0641,
"step": 15850
},
{
"epoch": 2.03,
"learning_rate": 8.578654126419095e-06,
"loss": 0.064,
"step": 15875
},
{
"epoch": 2.03,
"learning_rate": 8.570072990256342e-06,
"loss": 0.071,
"step": 15900
},
{
"epoch": 2.04,
"learning_rate": 8.561470348240068e-06,
"loss": 0.0681,
"step": 15925
},
{
"epoch": 2.04,
"learning_rate": 8.552846252191949e-06,
"loss": 0.0688,
"step": 15950
},
{
"epoch": 2.04,
"learning_rate": 8.544200754062894e-06,
"loss": 0.0687,
"step": 15975
},
{
"epoch": 2.04,
"learning_rate": 8.535533905932739e-06,
"loss": 0.0634,
"step": 16000
},
{
"epoch": 2.04,
"eval_cer": 5.690516052969069,
"eval_loss": 0.06705272942781448,
"eval_runtime": 3516.851,
"eval_samples_per_second": 2.586,
"eval_steps_per_second": 0.081,
"eval_wer": 10.69473523278875,
"step": 16000
},
{
"epoch": 2.05,
"learning_rate": 8.526845760009932e-06,
"loss": 0.0642,
"step": 16025
},
{
"epoch": 2.05,
"learning_rate": 8.518136368631216e-06,
"loss": 0.064,
"step": 16050
},
{
"epoch": 2.05,
"learning_rate": 8.509405784261315e-06,
"loss": 0.0665,
"step": 16075
},
{
"epoch": 2.06,
"learning_rate": 8.500654059492618e-06,
"loss": 0.0659,
"step": 16100
},
{
"epoch": 2.06,
"learning_rate": 8.491881247044866e-06,
"loss": 0.0692,
"step": 16125
},
{
"epoch": 2.06,
"learning_rate": 8.48308739976482e-06,
"loss": 0.0624,
"step": 16150
},
{
"epoch": 2.07,
"learning_rate": 8.474272570625967e-06,
"loss": 0.0628,
"step": 16175
},
{
"epoch": 2.07,
"learning_rate": 8.465436812728181e-06,
"loss": 0.0628,
"step": 16200
},
{
"epoch": 2.07,
"learning_rate": 8.456580179297402e-06,
"loss": 0.0608,
"step": 16225
},
{
"epoch": 2.08,
"learning_rate": 8.447702723685335e-06,
"loss": 0.0678,
"step": 16250
},
{
"epoch": 2.08,
"learning_rate": 8.438804499369108e-06,
"loss": 0.0639,
"step": 16275
},
{
"epoch": 2.08,
"learning_rate": 8.429885559950965e-06,
"loss": 0.0665,
"step": 16300
},
{
"epoch": 2.09,
"learning_rate": 8.420945959157926e-06,
"loss": 0.0668,
"step": 16325
},
{
"epoch": 2.09,
"learning_rate": 8.411985750841484e-06,
"loss": 0.069,
"step": 16350
},
{
"epoch": 2.09,
"learning_rate": 8.403004988977267e-06,
"loss": 0.0647,
"step": 16375
},
{
"epoch": 2.1,
"learning_rate": 8.39400372766471e-06,
"loss": 0.0647,
"step": 16400
},
{
"epoch": 2.1,
"learning_rate": 8.384982021126742e-06,
"loss": 0.0667,
"step": 16425
},
{
"epoch": 2.1,
"learning_rate": 8.375939923709453e-06,
"loss": 0.0624,
"step": 16450
},
{
"epoch": 2.11,
"learning_rate": 8.366877489881762e-06,
"loss": 0.0681,
"step": 16475
},
{
"epoch": 2.11,
"learning_rate": 8.357794774235094e-06,
"loss": 0.0652,
"step": 16500
},
{
"epoch": 2.11,
"learning_rate": 8.34869183148305e-06,
"loss": 0.0674,
"step": 16525
},
{
"epoch": 2.11,
"learning_rate": 8.339568716461082e-06,
"loss": 0.0678,
"step": 16550
},
{
"epoch": 2.12,
"learning_rate": 8.330425484126152e-06,
"loss": 0.0616,
"step": 16575
},
{
"epoch": 2.12,
"learning_rate": 8.32126218955641e-06,
"loss": 0.067,
"step": 16600
},
{
"epoch": 2.12,
"learning_rate": 8.31207888795086e-06,
"loss": 0.0697,
"step": 16625
},
{
"epoch": 2.13,
"learning_rate": 8.302875634629027e-06,
"loss": 0.0626,
"step": 16650
},
{
"epoch": 2.13,
"learning_rate": 8.29365248503062e-06,
"loss": 0.0681,
"step": 16675
},
{
"epoch": 2.13,
"learning_rate": 8.284409494715208e-06,
"loss": 0.0647,
"step": 16700
},
{
"epoch": 2.14,
"learning_rate": 8.275146719361872e-06,
"loss": 0.0656,
"step": 16725
},
{
"epoch": 2.14,
"learning_rate": 8.265864214768885e-06,
"loss": 0.0626,
"step": 16750
},
{
"epoch": 2.14,
"learning_rate": 8.256562036853358e-06,
"loss": 0.0654,
"step": 16775
},
{
"epoch": 2.15,
"learning_rate": 8.247240241650918e-06,
"loss": 0.0661,
"step": 16800
},
{
"epoch": 2.15,
"learning_rate": 8.237898885315368e-06,
"loss": 0.0656,
"step": 16825
},
{
"epoch": 2.15,
"learning_rate": 8.228538024118338e-06,
"loss": 0.066,
"step": 16850
},
{
"epoch": 2.16,
"learning_rate": 8.219157714448957e-06,
"loss": 0.0606,
"step": 16875
},
{
"epoch": 2.16,
"learning_rate": 8.209758012813515e-06,
"loss": 0.0597,
"step": 16900
},
{
"epoch": 2.16,
"learning_rate": 8.200338975835114e-06,
"loss": 0.0625,
"step": 16925
},
{
"epoch": 2.17,
"learning_rate": 8.190900660253327e-06,
"loss": 0.0632,
"step": 16950
},
{
"epoch": 2.17,
"learning_rate": 8.181443122923866e-06,
"loss": 0.0694,
"step": 16975
},
{
"epoch": 2.17,
"learning_rate": 8.171966420818227e-06,
"loss": 0.0618,
"step": 17000
},
{
"epoch": 2.17,
"eval_cer": 5.510712450983654,
"eval_loss": 0.0662335529923439,
"eval_runtime": 3548.5715,
"eval_samples_per_second": 2.562,
"eval_steps_per_second": 0.08,
"eval_wer": 10.292562566090941,
"step": 17000
},
{
"epoch": 2.18,
"learning_rate": 8.162470611023362e-06,
"loss": 0.0672,
"step": 17025
},
{
"epoch": 2.18,
"learning_rate": 8.15295575074132e-06,
"loss": 0.0649,
"step": 17050
},
{
"epoch": 2.18,
"learning_rate": 8.143421897288908e-06,
"loss": 0.0652,
"step": 17075
},
{
"epoch": 2.19,
"learning_rate": 8.133869108097349e-06,
"loss": 0.0621,
"step": 17100
},
{
"epoch": 2.19,
"learning_rate": 8.124297440711933e-06,
"loss": 0.0665,
"step": 17125
},
{
"epoch": 2.19,
"learning_rate": 8.11470695279167e-06,
"loss": 0.0652,
"step": 17150
},
{
"epoch": 2.19,
"learning_rate": 8.105097702108945e-06,
"loss": 0.0631,
"step": 17175
},
{
"epoch": 2.2,
"learning_rate": 8.095469746549172e-06,
"loss": 0.061,
"step": 17200
},
{
"epoch": 2.2,
"learning_rate": 8.08582314411043e-06,
"loss": 0.066,
"step": 17225
},
{
"epoch": 2.2,
"learning_rate": 8.076157952903134e-06,
"loss": 0.0627,
"step": 17250
},
{
"epoch": 2.21,
"learning_rate": 8.066474231149681e-06,
"loss": 0.0677,
"step": 17275
},
{
"epoch": 2.21,
"learning_rate": 8.056772037184083e-06,
"loss": 0.0633,
"step": 17300
},
{
"epoch": 2.21,
"learning_rate": 8.047051429451637e-06,
"loss": 0.0659,
"step": 17325
},
{
"epoch": 2.22,
"learning_rate": 8.037312466508555e-06,
"loss": 0.0626,
"step": 17350
},
{
"epoch": 2.22,
"learning_rate": 8.02755520702163e-06,
"loss": 0.0607,
"step": 17375
},
{
"epoch": 2.22,
"learning_rate": 8.017779709767857e-06,
"loss": 0.062,
"step": 17400
},
{
"epoch": 2.23,
"learning_rate": 8.00798603363411e-06,
"loss": 0.0665,
"step": 17425
},
{
"epoch": 2.23,
"learning_rate": 7.998174237616763e-06,
"loss": 0.0678,
"step": 17450
},
{
"epoch": 2.23,
"learning_rate": 7.98834438082134e-06,
"loss": 0.0708,
"step": 17475
},
{
"epoch": 2.24,
"learning_rate": 7.978496522462167e-06,
"loss": 0.0638,
"step": 17500
},
{
"epoch": 2.24,
"learning_rate": 7.96863072186201e-06,
"loss": 0.0612,
"step": 17525
},
{
"epoch": 2.24,
"learning_rate": 7.958747038451715e-06,
"loss": 0.0667,
"step": 17550
},
{
"epoch": 2.25,
"learning_rate": 7.948845531769855e-06,
"loss": 0.0649,
"step": 17575
},
{
"epoch": 2.25,
"learning_rate": 7.938926261462366e-06,
"loss": 0.0683,
"step": 17600
},
{
"epoch": 2.25,
"learning_rate": 7.928989287282195e-06,
"loss": 0.0644,
"step": 17625
},
{
"epoch": 2.26,
"learning_rate": 7.919034669088933e-06,
"loss": 0.0682,
"step": 17650
},
{
"epoch": 2.26,
"learning_rate": 7.909062466848456e-06,
"loss": 0.0579,
"step": 17675
},
{
"epoch": 2.26,
"learning_rate": 7.89907274063257e-06,
"loss": 0.065,
"step": 17700
},
{
"epoch": 2.27,
"learning_rate": 7.889065550618637e-06,
"loss": 0.0653,
"step": 17725
},
{
"epoch": 2.27,
"learning_rate": 7.879040957089229e-06,
"loss": 0.0679,
"step": 17750
},
{
"epoch": 2.27,
"learning_rate": 7.868999020431742e-06,
"loss": 0.0629,
"step": 17775
},
{
"epoch": 2.27,
"learning_rate": 7.858939801138061e-06,
"loss": 0.064,
"step": 17800
},
{
"epoch": 2.28,
"learning_rate": 7.848863359804172e-06,
"loss": 0.068,
"step": 17825
},
{
"epoch": 2.28,
"learning_rate": 7.838769757129804e-06,
"loss": 0.0635,
"step": 17850
},
{
"epoch": 2.28,
"learning_rate": 7.828659053918067e-06,
"loss": 0.0641,
"step": 17875
},
{
"epoch": 2.29,
"learning_rate": 7.818531311075084e-06,
"loss": 0.0674,
"step": 17900
},
{
"epoch": 2.29,
"learning_rate": 7.808386589609624e-06,
"loss": 0.0608,
"step": 17925
},
{
"epoch": 2.29,
"learning_rate": 7.79822495063273e-06,
"loss": 0.0654,
"step": 17950
},
{
"epoch": 2.3,
"learning_rate": 7.78804645535736e-06,
"loss": 0.0661,
"step": 17975
},
{
"epoch": 2.3,
"learning_rate": 7.777851165098012e-06,
"loss": 0.0679,
"step": 18000
},
{
"epoch": 2.3,
"eval_cer": 5.494820251835918,
"eval_loss": 0.06429608911275864,
"eval_runtime": 3561.7203,
"eval_samples_per_second": 2.553,
"eval_steps_per_second": 0.08,
"eval_wer": 10.179172806842455,
"step": 18000
},
{
"epoch": 2.3,
"learning_rate": 7.76763914127035e-06,
"loss": 0.0625,
"step": 18025
},
{
"epoch": 2.31,
"learning_rate": 7.757410445390847e-06,
"loss": 0.0599,
"step": 18050
},
{
"epoch": 2.31,
"learning_rate": 7.747165139076405e-06,
"loss": 0.0602,
"step": 18075
},
{
"epoch": 2.31,
"learning_rate": 7.736903284043985e-06,
"loss": 0.0621,
"step": 18100
},
{
"epoch": 2.32,
"learning_rate": 7.726624942110233e-06,
"loss": 0.0564,
"step": 18125
},
{
"epoch": 2.32,
"learning_rate": 7.716330175191118e-06,
"loss": 0.0678,
"step": 18150
},
{
"epoch": 2.32,
"learning_rate": 7.706019045301549e-06,
"loss": 0.058,
"step": 18175
},
{
"epoch": 2.33,
"learning_rate": 7.695691614555002e-06,
"loss": 0.0625,
"step": 18200
},
{
"epoch": 2.33,
"learning_rate": 7.68534794516315e-06,
"loss": 0.0628,
"step": 18225
},
{
"epoch": 2.33,
"learning_rate": 7.674988099435487e-06,
"loss": 0.0656,
"step": 18250
},
{
"epoch": 2.34,
"learning_rate": 7.664612139778951e-06,
"loss": 0.0641,
"step": 18275
},
{
"epoch": 2.34,
"learning_rate": 7.654220128697547e-06,
"loss": 0.0604,
"step": 18300
},
{
"epoch": 2.34,
"learning_rate": 7.64381212879198e-06,
"loss": 0.059,
"step": 18325
},
{
"epoch": 2.35,
"learning_rate": 7.633388202759262e-06,
"loss": 0.068,
"step": 18350
},
{
"epoch": 2.35,
"learning_rate": 7.6229484133923445e-06,
"loss": 0.0624,
"step": 18375
},
{
"epoch": 2.35,
"learning_rate": 7.612492823579744e-06,
"loss": 0.0671,
"step": 18400
},
{
"epoch": 2.35,
"learning_rate": 7.602021496305154e-06,
"loss": 0.0576,
"step": 18425
},
{
"epoch": 2.36,
"learning_rate": 7.591534494647066e-06,
"loss": 0.0609,
"step": 18450
},
{
"epoch": 2.36,
"learning_rate": 7.5810318817784e-06,
"loss": 0.0634,
"step": 18475
},
{
"epoch": 2.36,
"learning_rate": 7.570513720966108e-06,
"loss": 0.0606,
"step": 18500
},
{
"epoch": 2.37,
"learning_rate": 7.559980075570813e-06,
"loss": 0.0633,
"step": 18525
},
{
"epoch": 2.37,
"learning_rate": 7.549431009046404e-06,
"loss": 0.0625,
"step": 18550
},
{
"epoch": 2.37,
"learning_rate": 7.538866584939673e-06,
"loss": 0.0647,
"step": 18575
},
{
"epoch": 2.38,
"learning_rate": 7.528286866889924e-06,
"loss": 0.0594,
"step": 18600
},
{
"epoch": 2.38,
"learning_rate": 7.517691918628589e-06,
"loss": 0.062,
"step": 18625
},
{
"epoch": 2.38,
"learning_rate": 7.5070818039788455e-06,
"loss": 0.0638,
"step": 18650
},
{
"epoch": 2.39,
"learning_rate": 7.496456586855236e-06,
"loss": 0.0626,
"step": 18675
},
{
"epoch": 2.39,
"learning_rate": 7.485816331263273e-06,
"loss": 0.0613,
"step": 18700
},
{
"epoch": 2.39,
"learning_rate": 7.475161101299065e-06,
"loss": 0.0585,
"step": 18725
},
{
"epoch": 2.4,
"learning_rate": 7.464490961148921e-06,
"loss": 0.0602,
"step": 18750
},
{
"epoch": 2.4,
"learning_rate": 7.453805975088971e-06,
"loss": 0.0651,
"step": 18775
},
{
"epoch": 2.4,
"learning_rate": 7.443106207484776e-06,
"loss": 0.0611,
"step": 18800
},
{
"epoch": 2.41,
"learning_rate": 7.432391722790934e-06,
"loss": 0.0655,
"step": 18825
},
{
"epoch": 2.41,
"learning_rate": 7.421662585550707e-06,
"loss": 0.0607,
"step": 18850
},
{
"epoch": 2.41,
"learning_rate": 7.410918860395615e-06,
"loss": 0.0684,
"step": 18875
},
{
"epoch": 2.42,
"learning_rate": 7.400160612045057e-06,
"loss": 0.0621,
"step": 18900
},
{
"epoch": 2.42,
"learning_rate": 7.389387905305918e-06,
"loss": 0.0654,
"step": 18925
},
{
"epoch": 2.42,
"learning_rate": 7.378600805072186e-06,
"loss": 0.0592,
"step": 18950
},
{
"epoch": 2.42,
"learning_rate": 7.367799376324542e-06,
"loss": 0.0654,
"step": 18975
},
{
"epoch": 2.43,
"learning_rate": 7.3569836841299905e-06,
"loss": 0.0589,
"step": 19000
},
{
"epoch": 2.43,
"eval_cer": 5.520136894664287,
"eval_loss": 0.06469350308179855,
"eval_runtime": 3562.4947,
"eval_samples_per_second": 2.552,
"eval_steps_per_second": 0.08,
"eval_wer": 10.188091102513685,
"step": 19000
},
{
"epoch": 2.43,
"learning_rate": 7.346153793641453e-06,
"loss": 0.0663,
"step": 19025
},
{
"epoch": 2.43,
"learning_rate": 7.335309770097383e-06,
"loss": 0.0634,
"step": 19050
},
{
"epoch": 2.44,
"learning_rate": 7.324451678821374e-06,
"loss": 0.0628,
"step": 19075
},
{
"epoch": 2.44,
"learning_rate": 7.313579585221752e-06,
"loss": 0.0575,
"step": 19100
},
{
"epoch": 2.44,
"learning_rate": 7.3026935547912004e-06,
"loss": 0.0613,
"step": 19125
},
{
"epoch": 2.45,
"learning_rate": 7.291793653106357e-06,
"loss": 0.0588,
"step": 19150
},
{
"epoch": 2.45,
"learning_rate": 7.280879945827415e-06,
"loss": 0.061,
"step": 19175
},
{
"epoch": 2.45,
"learning_rate": 7.269952498697734e-06,
"loss": 0.0617,
"step": 19200
},
{
"epoch": 2.46,
"learning_rate": 7.2590113775434415e-06,
"loss": 0.0625,
"step": 19225
},
{
"epoch": 2.46,
"learning_rate": 7.248056648273034e-06,
"loss": 0.0575,
"step": 19250
},
{
"epoch": 2.46,
"learning_rate": 7.237088376876984e-06,
"loss": 0.0604,
"step": 19275
},
{
"epoch": 2.47,
"learning_rate": 7.226106629427342e-06,
"loss": 0.061,
"step": 19300
},
{
"epoch": 2.47,
"learning_rate": 7.215111472077334e-06,
"loss": 0.0593,
"step": 19325
},
{
"epoch": 2.47,
"learning_rate": 7.204102971060971e-06,
"loss": 0.0591,
"step": 19350
},
{
"epoch": 2.48,
"learning_rate": 7.193081192692639e-06,
"loss": 0.0592,
"step": 19375
},
{
"epoch": 2.48,
"learning_rate": 7.18204620336671e-06,
"loss": 0.0619,
"step": 19400
},
{
"epoch": 2.48,
"learning_rate": 7.17099806955714e-06,
"loss": 0.0604,
"step": 19425
},
{
"epoch": 2.49,
"learning_rate": 7.15993685781706e-06,
"loss": 0.0595,
"step": 19450
},
{
"epoch": 2.49,
"learning_rate": 7.148862634778385e-06,
"loss": 0.0604,
"step": 19475
},
{
"epoch": 2.49,
"learning_rate": 7.137775467151411e-06,
"loss": 0.0558,
"step": 19500
},
{
"epoch": 2.5,
"learning_rate": 7.1266754217244075e-06,
"loss": 0.0565,
"step": 19525
},
{
"epoch": 2.5,
"learning_rate": 7.115562565363221e-06,
"loss": 0.0611,
"step": 19550
},
{
"epoch": 2.5,
"learning_rate": 7.10443696501087e-06,
"loss": 0.0663,
"step": 19575
},
{
"epoch": 2.5,
"learning_rate": 7.093298687687141e-06,
"loss": 0.0624,
"step": 19600
},
{
"epoch": 2.51,
"learning_rate": 7.0821478004881875e-06,
"loss": 0.0612,
"step": 19625
},
{
"epoch": 2.51,
"learning_rate": 7.070984370586119e-06,
"loss": 0.0632,
"step": 19650
},
{
"epoch": 2.51,
"learning_rate": 7.059808465228609e-06,
"loss": 0.0531,
"step": 19675
},
{
"epoch": 2.52,
"learning_rate": 7.048620151738478e-06,
"loss": 0.062,
"step": 19700
},
{
"epoch": 2.52,
"learning_rate": 7.037419497513289e-06,
"loss": 0.0615,
"step": 19725
},
{
"epoch": 2.52,
"learning_rate": 7.026206570024949e-06,
"loss": 0.0606,
"step": 19750
},
{
"epoch": 2.53,
"learning_rate": 7.0149814368193e-06,
"loss": 0.0585,
"step": 19775
},
{
"epoch": 2.53,
"learning_rate": 7.0037441655157045e-06,
"loss": 0.0609,
"step": 19800
},
{
"epoch": 2.53,
"learning_rate": 6.992494823806651e-06,
"loss": 0.0568,
"step": 19825
},
{
"epoch": 2.54,
"learning_rate": 6.9812334794573285e-06,
"loss": 0.0603,
"step": 19850
},
{
"epoch": 2.54,
"learning_rate": 6.969960200305242e-06,
"loss": 0.0616,
"step": 19875
},
{
"epoch": 2.54,
"learning_rate": 6.95867505425978e-06,
"loss": 0.0561,
"step": 19900
},
{
"epoch": 2.55,
"learning_rate": 6.947378109301825e-06,
"loss": 0.0626,
"step": 19925
},
{
"epoch": 2.55,
"learning_rate": 6.936069433483329e-06,
"loss": 0.0599,
"step": 19950
},
{
"epoch": 2.55,
"learning_rate": 6.924749094926913e-06,
"loss": 0.0608,
"step": 19975
},
{
"epoch": 2.56,
"learning_rate": 6.913417161825449e-06,
"loss": 0.0623,
"step": 20000
},
{
"epoch": 2.56,
"eval_cer": 5.273068635821017,
"eval_loss": 0.06334798038005829,
"eval_runtime": 3564.0622,
"eval_samples_per_second": 2.551,
"eval_steps_per_second": 0.08,
"eval_wer": 9.844949059544488,
"step": 20000
},
{
"epoch": 2.56,
"learning_rate": 6.902073702441662e-06,
"loss": 0.0589,
"step": 20025
},
{
"epoch": 2.56,
"learning_rate": 6.8907187851077026e-06,
"loss": 0.0597,
"step": 20050
},
{
"epoch": 2.57,
"learning_rate": 6.879352478224749e-06,
"loss": 0.0581,
"step": 20075
},
{
"epoch": 2.57,
"learning_rate": 6.867974850262582e-06,
"loss": 0.0598,
"step": 20100
},
{
"epoch": 2.57,
"learning_rate": 6.856585969759189e-06,
"loss": 0.06,
"step": 20125
},
{
"epoch": 2.58,
"learning_rate": 6.845185905320333e-06,
"loss": 0.0607,
"step": 20150
},
{
"epoch": 2.58,
"learning_rate": 6.833774725619154e-06,
"loss": 0.062,
"step": 20175
},
{
"epoch": 2.58,
"learning_rate": 6.822352499395751e-06,
"loss": 0.0636,
"step": 20200
},
{
"epoch": 2.58,
"learning_rate": 6.8109192954567546e-06,
"loss": 0.0638,
"step": 20225
},
{
"epoch": 2.59,
"learning_rate": 6.799475182674942e-06,
"loss": 0.0599,
"step": 20250
},
{
"epoch": 2.59,
"learning_rate": 6.7880202299887895e-06,
"loss": 0.0675,
"step": 20275
},
{
"epoch": 2.59,
"learning_rate": 6.776554506402081e-06,
"loss": 0.055,
"step": 20300
},
{
"epoch": 2.6,
"learning_rate": 6.7650780809834806e-06,
"loss": 0.0645,
"step": 20325
},
{
"epoch": 2.6,
"learning_rate": 6.753591022866117e-06,
"loss": 0.0573,
"step": 20350
},
{
"epoch": 2.6,
"learning_rate": 6.742093401247173e-06,
"loss": 0.0593,
"step": 20375
},
{
"epoch": 2.61,
"learning_rate": 6.730585285387465e-06,
"loss": 0.0606,
"step": 20400
},
{
"epoch": 2.61,
"learning_rate": 6.719066744611024e-06,
"loss": 0.0624,
"step": 20425
},
{
"epoch": 2.61,
"learning_rate": 6.707537848304682e-06,
"loss": 0.062,
"step": 20450
},
{
"epoch": 2.62,
"learning_rate": 6.69599866591765e-06,
"loss": 0.0612,
"step": 20475
},
{
"epoch": 2.62,
"learning_rate": 6.684449266961101e-06,
"loss": 0.0616,
"step": 20500
},
{
"epoch": 2.62,
"learning_rate": 6.672889721007756e-06,
"loss": 0.055,
"step": 20525
},
{
"epoch": 2.63,
"learning_rate": 6.661320097691454e-06,
"loss": 0.0574,
"step": 20550
},
{
"epoch": 2.63,
"learning_rate": 6.6497404667067474e-06,
"loss": 0.059,
"step": 20575
},
{
"epoch": 2.63,
"learning_rate": 6.638150897808469e-06,
"loss": 0.0682,
"step": 20600
},
{
"epoch": 2.64,
"learning_rate": 6.626551460811316e-06,
"loss": 0.0601,
"step": 20625
},
{
"epoch": 2.64,
"learning_rate": 6.614942225589432e-06,
"loss": 0.0556,
"step": 20650
},
{
"epoch": 2.64,
"learning_rate": 6.603323262075987e-06,
"loss": 0.0576,
"step": 20675
},
{
"epoch": 2.65,
"learning_rate": 6.591694640262749e-06,
"loss": 0.0586,
"step": 20700
},
{
"epoch": 2.65,
"learning_rate": 6.58005643019967e-06,
"loss": 0.0631,
"step": 20725
},
{
"epoch": 2.65,
"learning_rate": 6.568408701994459e-06,
"loss": 0.0586,
"step": 20750
},
{
"epoch": 2.65,
"learning_rate": 6.5567515258121605e-06,
"loss": 0.059,
"step": 20775
},
{
"epoch": 2.66,
"learning_rate": 6.545084971874738e-06,
"loss": 0.0579,
"step": 20800
},
{
"epoch": 2.66,
"learning_rate": 6.533409110460641e-06,
"loss": 0.0578,
"step": 20825
},
{
"epoch": 2.66,
"learning_rate": 6.521724011904387e-06,
"loss": 0.0598,
"step": 20850
},
{
"epoch": 2.67,
"learning_rate": 6.510029746596141e-06,
"loss": 0.0585,
"step": 20875
},
{
"epoch": 2.67,
"learning_rate": 6.4983263849812835e-06,
"loss": 0.0565,
"step": 20900
},
{
"epoch": 2.67,
"learning_rate": 6.486613997559993e-06,
"loss": 0.0629,
"step": 20925
},
{
"epoch": 2.68,
"learning_rate": 6.474892654886819e-06,
"loss": 0.0608,
"step": 20950
},
{
"epoch": 2.68,
"learning_rate": 6.4631624275702535e-06,
"loss": 0.0572,
"step": 20975
},
{
"epoch": 2.68,
"learning_rate": 6.451423386272312e-06,
"loss": 0.0558,
"step": 21000
},
{
"epoch": 2.68,
"eval_cer": 5.421087839510964,
"eval_loss": 0.062320612370967865,
"eval_runtime": 3611.6505,
"eval_samples_per_second": 2.518,
"eval_steps_per_second": 0.079,
"eval_wer": 10.026712418939063,
"step": 21000
},
{
"epoch": 2.69,
"learning_rate": 6.439675601708106e-06,
"loss": 0.0565,
"step": 21025
},
{
"epoch": 2.69,
"learning_rate": 6.427919144645411e-06,
"loss": 0.056,
"step": 21050
},
{
"epoch": 2.69,
"learning_rate": 6.4161540859042524e-06,
"loss": 0.0551,
"step": 21075
},
{
"epoch": 2.7,
"learning_rate": 6.4043804963564616e-06,
"loss": 0.0577,
"step": 21100
},
{
"epoch": 2.7,
"learning_rate": 6.392598446925266e-06,
"loss": 0.059,
"step": 21125
},
{
"epoch": 2.7,
"learning_rate": 6.3808080085848544e-06,
"loss": 0.0605,
"step": 21150
},
{
"epoch": 2.71,
"learning_rate": 6.369009252359947e-06,
"loss": 0.06,
"step": 21175
},
{
"epoch": 2.71,
"learning_rate": 6.3572022493253715e-06,
"loss": 0.0623,
"step": 21200
},
{
"epoch": 2.71,
"learning_rate": 6.3453870706056355e-06,
"loss": 0.0542,
"step": 21225
},
{
"epoch": 2.72,
"learning_rate": 6.333563787374493e-06,
"loss": 0.0562,
"step": 21250
},
{
"epoch": 2.72,
"learning_rate": 6.321732470854521e-06,
"loss": 0.0604,
"step": 21275
},
{
"epoch": 2.72,
"learning_rate": 6.309893192316687e-06,
"loss": 0.0579,
"step": 21300
},
{
"epoch": 2.73,
"learning_rate": 6.298046023079927e-06,
"loss": 0.0577,
"step": 21325
},
{
"epoch": 2.73,
"learning_rate": 6.2861910345107e-06,
"loss": 0.0542,
"step": 21350
},
{
"epoch": 2.73,
"learning_rate": 6.274328298022574e-06,
"loss": 0.0587,
"step": 21375
},
{
"epoch": 2.73,
"learning_rate": 6.26245788507579e-06,
"loss": 0.0603,
"step": 21400
},
{
"epoch": 2.74,
"learning_rate": 6.250579867176832e-06,
"loss": 0.0561,
"step": 21425
},
{
"epoch": 2.74,
"learning_rate": 6.238694315877994e-06,
"loss": 0.056,
"step": 21450
},
{
"epoch": 2.74,
"learning_rate": 6.226801302776947e-06,
"loss": 0.0601,
"step": 21475
},
{
"epoch": 2.75,
"learning_rate": 6.21490089951632e-06,
"loss": 0.0536,
"step": 21500
},
{
"epoch": 2.75,
"learning_rate": 6.202993177783253e-06,
"loss": 0.061,
"step": 21525
},
{
"epoch": 2.75,
"learning_rate": 6.191078209308974e-06,
"loss": 0.0556,
"step": 21550
},
{
"epoch": 2.76,
"learning_rate": 6.179156065868365e-06,
"loss": 0.0609,
"step": 21575
},
{
"epoch": 2.76,
"learning_rate": 6.1672268192795285e-06,
"loss": 0.0517,
"step": 21600
},
{
"epoch": 2.76,
"learning_rate": 6.155290541403357e-06,
"loss": 0.0562,
"step": 21625
},
{
"epoch": 2.77,
"learning_rate": 6.143347304143098e-06,
"loss": 0.0596,
"step": 21650
},
{
"epoch": 2.77,
"learning_rate": 6.1313971794439235e-06,
"loss": 0.0608,
"step": 21675
},
{
"epoch": 2.77,
"learning_rate": 6.119440239292493e-06,
"loss": 0.0576,
"step": 21700
},
{
"epoch": 2.78,
"learning_rate": 6.107476555716523e-06,
"loss": 0.0578,
"step": 21725
},
{
"epoch": 2.78,
"learning_rate": 6.095506200784349e-06,
"loss": 0.0605,
"step": 21750
},
{
"epoch": 2.78,
"learning_rate": 6.0835292466045e-06,
"loss": 0.0561,
"step": 21775
},
{
"epoch": 2.79,
"learning_rate": 6.071545765325254e-06,
"loss": 0.0547,
"step": 21800
},
{
"epoch": 2.79,
"learning_rate": 6.059555829134208e-06,
"loss": 0.0505,
"step": 21825
},
{
"epoch": 2.79,
"learning_rate": 6.0475595102578455e-06,
"loss": 0.0549,
"step": 21850
},
{
"epoch": 2.8,
"learning_rate": 6.035556880961093e-06,
"loss": 0.0559,
"step": 21875
},
{
"epoch": 2.8,
"learning_rate": 6.023548013546899e-06,
"loss": 0.0587,
"step": 21900
},
{
"epoch": 2.8,
"learning_rate": 6.011532980355782e-06,
"loss": 0.0633,
"step": 21925
},
{
"epoch": 2.81,
"learning_rate": 5.99951185376541e-06,
"loss": 0.0571,
"step": 21950
},
{
"epoch": 2.81,
"learning_rate": 5.98748470619015e-06,
"loss": 0.055,
"step": 21975
},
{
"epoch": 2.81,
"learning_rate": 5.975451610080643e-06,
"loss": 0.0564,
"step": 22000
},
{
"epoch": 2.81,
"eval_cer": 5.455274546979927,
"eval_loss": 0.06174223870038986,
"eval_runtime": 3638.3404,
"eval_samples_per_second": 2.499,
"eval_steps_per_second": 0.078,
"eval_wer": 9.98934051326915,
"step": 22000
},
{
"epoch": 2.81,
"learning_rate": 5.963412637923361e-06,
"loss": 0.0621,
"step": 22025
},
{
"epoch": 2.82,
"learning_rate": 5.95136786224018e-06,
"loss": 0.057,
"step": 22050
},
{
"epoch": 2.82,
"learning_rate": 5.939317355587927e-06,
"loss": 0.0539,
"step": 22075
},
{
"epoch": 2.82,
"learning_rate": 5.927261190557955e-06,
"loss": 0.0592,
"step": 22100
},
{
"epoch": 2.83,
"learning_rate": 5.915199439775706e-06,
"loss": 0.054,
"step": 22125
},
{
"epoch": 2.83,
"learning_rate": 5.903132175900264e-06,
"loss": 0.0579,
"step": 22150
},
{
"epoch": 2.83,
"learning_rate": 5.89105947162393e-06,
"loss": 0.0589,
"step": 22175
},
{
"epoch": 2.84,
"learning_rate": 5.878981399671774e-06,
"loss": 0.0589,
"step": 22200
},
{
"epoch": 2.84,
"learning_rate": 5.8668980328011985e-06,
"loss": 0.0558,
"step": 22225
},
{
"epoch": 2.84,
"learning_rate": 5.8548094438015065e-06,
"loss": 0.0588,
"step": 22250
},
{
"epoch": 2.85,
"learning_rate": 5.8427157054934565e-06,
"loss": 0.0565,
"step": 22275
},
{
"epoch": 2.85,
"learning_rate": 5.830616890728828e-06,
"loss": 0.0563,
"step": 22300
},
{
"epoch": 2.85,
"learning_rate": 5.818513072389978e-06,
"loss": 0.0586,
"step": 22325
},
{
"epoch": 2.86,
"learning_rate": 5.806404323389403e-06,
"loss": 0.0573,
"step": 22350
},
{
"epoch": 2.86,
"learning_rate": 5.794290716669307e-06,
"loss": 0.0558,
"step": 22375
},
{
"epoch": 2.86,
"learning_rate": 5.782172325201155e-06,
"loss": 0.0496,
"step": 22400
},
{
"epoch": 2.87,
"learning_rate": 5.770049221985231e-06,
"loss": 0.0591,
"step": 22425
},
{
"epoch": 2.87,
"learning_rate": 5.757921480050206e-06,
"loss": 0.0578,
"step": 22450
},
{
"epoch": 2.87,
"learning_rate": 5.745789172452694e-06,
"loss": 0.0529,
"step": 22475
},
{
"epoch": 2.88,
"learning_rate": 5.733652372276809e-06,
"loss": 0.0565,
"step": 22500
},
{
"epoch": 2.88,
"learning_rate": 5.721511152633734e-06,
"loss": 0.0532,
"step": 22525
},
{
"epoch": 2.88,
"learning_rate": 5.709365586661266e-06,
"loss": 0.0588,
"step": 22550
},
{
"epoch": 2.88,
"learning_rate": 5.697215747523394e-06,
"loss": 0.0531,
"step": 22575
},
{
"epoch": 2.89,
"learning_rate": 5.6850617084098416e-06,
"loss": 0.0566,
"step": 22600
},
{
"epoch": 2.89,
"learning_rate": 5.672903542535631e-06,
"loss": 0.0556,
"step": 22625
},
{
"epoch": 2.89,
"learning_rate": 5.660741323140651e-06,
"loss": 0.0556,
"step": 22650
},
{
"epoch": 2.9,
"learning_rate": 5.648575123489203e-06,
"loss": 0.057,
"step": 22675
},
{
"epoch": 2.9,
"learning_rate": 5.636405016869567e-06,
"loss": 0.0531,
"step": 22700
},
{
"epoch": 2.9,
"learning_rate": 5.624231076593556e-06,
"loss": 0.054,
"step": 22725
},
{
"epoch": 2.91,
"learning_rate": 5.612053375996082e-06,
"loss": 0.0583,
"step": 22750
},
{
"epoch": 2.91,
"learning_rate": 5.5998719884347015e-06,
"loss": 0.0548,
"step": 22775
},
{
"epoch": 2.91,
"learning_rate": 5.587686987289189e-06,
"loss": 0.0527,
"step": 22800
},
{
"epoch": 2.92,
"learning_rate": 5.57549844596108e-06,
"loss": 0.0578,
"step": 22825
},
{
"epoch": 2.92,
"learning_rate": 5.563306437873239e-06,
"loss": 0.0536,
"step": 22850
},
{
"epoch": 2.92,
"learning_rate": 5.551111036469416e-06,
"loss": 0.0528,
"step": 22875
},
{
"epoch": 2.93,
"learning_rate": 5.5389123152137965e-06,
"loss": 0.0536,
"step": 22900
},
{
"epoch": 2.93,
"learning_rate": 5.52671034759057e-06,
"loss": 0.0549,
"step": 22925
},
{
"epoch": 2.93,
"learning_rate": 5.514505207103482e-06,
"loss": 0.0604,
"step": 22950
},
{
"epoch": 2.94,
"learning_rate": 5.5022969672753835e-06,
"loss": 0.0586,
"step": 22975
},
{
"epoch": 2.94,
"learning_rate": 5.490085701647805e-06,
"loss": 0.0552,
"step": 23000
},
{
"epoch": 2.94,
"eval_cer": 5.385977166975271,
"eval_loss": 0.06067069619894028,
"eval_runtime": 3681.7154,
"eval_samples_per_second": 2.47,
"eval_steps_per_second": 0.077,
"eval_wer": 9.777849501637144,
"step": 23000
},
{
"epoch": 2.94,
"learning_rate": 5.477871483780497e-06,
"loss": 0.055,
"step": 23025
},
{
"epoch": 2.95,
"learning_rate": 5.4656543872509994e-06,
"loss": 0.0528,
"step": 23050
},
{
"epoch": 2.95,
"learning_rate": 5.453434485654192e-06,
"loss": 0.0594,
"step": 23075
},
{
"epoch": 2.95,
"learning_rate": 5.441211852601849e-06,
"loss": 0.0586,
"step": 23100
},
{
"epoch": 2.96,
"learning_rate": 5.4289865617222005e-06,
"loss": 0.0556,
"step": 23125
},
{
"epoch": 2.96,
"learning_rate": 5.416758686659488e-06,
"loss": 0.0508,
"step": 23150
},
{
"epoch": 2.96,
"learning_rate": 5.40452830107352e-06,
"loss": 0.0558,
"step": 23175
},
{
"epoch": 2.96,
"learning_rate": 5.392295478639226e-06,
"loss": 0.0548,
"step": 23200
},
{
"epoch": 2.97,
"learning_rate": 5.380060293046216e-06,
"loss": 0.0522,
"step": 23225
},
{
"epoch": 2.97,
"learning_rate": 5.367822817998338e-06,
"loss": 0.0534,
"step": 23250
},
{
"epoch": 2.97,
"learning_rate": 5.355583127213227e-06,
"loss": 0.0544,
"step": 23275
},
{
"epoch": 2.98,
"learning_rate": 5.343341294421868e-06,
"loss": 0.0466,
"step": 23300
},
{
"epoch": 2.98,
"learning_rate": 5.331097393368153e-06,
"loss": 0.05,
"step": 23325
},
{
"epoch": 2.98,
"learning_rate": 5.318851497808424e-06,
"loss": 0.049,
"step": 23350
},
{
"epoch": 2.99,
"learning_rate": 5.306603681511043e-06,
"loss": 0.0556,
"step": 23375
},
{
"epoch": 2.99,
"learning_rate": 5.294354018255945e-06,
"loss": 0.0519,
"step": 23400
},
{
"epoch": 2.99,
"learning_rate": 5.282102581834188e-06,
"loss": 0.0522,
"step": 23425
},
{
"epoch": 3.0,
"learning_rate": 5.26984944604751e-06,
"loss": 0.0566,
"step": 23450
},
{
"epoch": 3.0,
"learning_rate": 5.2575946847078875e-06,
"loss": 0.053,
"step": 23475
},
{
"epoch": 3.0,
"learning_rate": 5.245338371637091e-06,
"loss": 0.0439,
"step": 23500
},
{
"epoch": 3.01,
"learning_rate": 5.233080580666234e-06,
"loss": 0.0448,
"step": 23525
},
{
"epoch": 3.01,
"learning_rate": 5.220821385635337e-06,
"loss": 0.0472,
"step": 23550
},
{
"epoch": 3.01,
"learning_rate": 5.208560860392879e-06,
"loss": 0.0425,
"step": 23575
},
{
"epoch": 3.02,
"learning_rate": 5.1962990787953436e-06,
"loss": 0.0398,
"step": 23600
},
{
"epoch": 3.02,
"learning_rate": 5.184036114706795e-06,
"loss": 0.0411,
"step": 23625
},
{
"epoch": 3.02,
"learning_rate": 5.171772041998412e-06,
"loss": 0.0419,
"step": 23650
},
{
"epoch": 3.03,
"learning_rate": 5.159506934548054e-06,
"loss": 0.0427,
"step": 23675
},
{
"epoch": 3.03,
"learning_rate": 5.147240866239817e-06,
"loss": 0.0437,
"step": 23700
},
{
"epoch": 3.03,
"learning_rate": 5.1349739109635775e-06,
"loss": 0.0452,
"step": 23725
},
{
"epoch": 3.04,
"learning_rate": 5.122706142614562e-06,
"loss": 0.0382,
"step": 23750
},
{
"epoch": 3.04,
"learning_rate": 5.110437635092892e-06,
"loss": 0.044,
"step": 23775
},
{
"epoch": 3.04,
"learning_rate": 5.098168462303141e-06,
"loss": 0.0435,
"step": 23800
},
{
"epoch": 3.04,
"learning_rate": 5.085898698153895e-06,
"loss": 0.0423,
"step": 23825
},
{
"epoch": 3.05,
"learning_rate": 5.073628416557293e-06,
"loss": 0.0398,
"step": 23850
},
{
"epoch": 3.05,
"learning_rate": 5.0613576914286e-06,
"loss": 0.0438,
"step": 23875
},
{
"epoch": 3.05,
"learning_rate": 5.049086596685749e-06,
"loss": 0.0422,
"step": 23900
},
{
"epoch": 3.06,
"learning_rate": 5.036815206248898e-06,
"loss": 0.039,
"step": 23925
},
{
"epoch": 3.06,
"learning_rate": 5.024543594039991e-06,
"loss": 0.0409,
"step": 23950
},
{
"epoch": 3.06,
"learning_rate": 5.0122718339823025e-06,
"loss": 0.0435,
"step": 23975
},
{
"epoch": 3.07,
"learning_rate": 5e-06,
"loss": 0.0403,
"step": 24000
},
{
"epoch": 3.07,
"eval_cer": 5.729692171798368,
"eval_loss": 0.06206570565700531,
"eval_runtime": 3767.6952,
"eval_samples_per_second": 2.413,
"eval_steps_per_second": 0.076,
"eval_wer": 10.038178799087785,
"step": 24000
},
{
"epoch": 3.07,
"learning_rate": 4.987728166017699e-06,
"loss": 0.0429,
"step": 24025
},
{
"epoch": 3.07,
"learning_rate": 4.97545640596001e-06,
"loss": 0.0476,
"step": 24050
},
{
"epoch": 3.08,
"learning_rate": 4.963184793751103e-06,
"loss": 0.0446,
"step": 24075
},
{
"epoch": 3.08,
"learning_rate": 4.9509134033142525e-06,
"loss": 0.0461,
"step": 24100
},
{
"epoch": 3.08,
"learning_rate": 4.938642308571401e-06,
"loss": 0.0438,
"step": 24125
},
{
"epoch": 3.09,
"learning_rate": 4.926371583442709e-06,
"loss": 0.0463,
"step": 24150
},
{
"epoch": 3.09,
"learning_rate": 4.914101301846107e-06,
"loss": 0.0416,
"step": 24175
},
{
"epoch": 3.09,
"learning_rate": 4.90183153769686e-06,
"loss": 0.0421,
"step": 24200
},
{
"epoch": 3.1,
"learning_rate": 4.889562364907108e-06,
"loss": 0.0408,
"step": 24225
},
{
"epoch": 3.1,
"learning_rate": 4.87729385738544e-06,
"loss": 0.0395,
"step": 24250
},
{
"epoch": 3.1,
"learning_rate": 4.865026089036425e-06,
"loss": 0.0411,
"step": 24275
},
{
"epoch": 3.11,
"learning_rate": 4.852759133760184e-06,
"loss": 0.0424,
"step": 24300
},
{
"epoch": 3.11,
"learning_rate": 4.840493065451947e-06,
"loss": 0.0418,
"step": 24325
},
{
"epoch": 3.11,
"learning_rate": 4.828227958001589e-06,
"loss": 0.0441,
"step": 24350
},
{
"epoch": 3.11,
"learning_rate": 4.815963885293206e-06,
"loss": 0.0421,
"step": 24375
},
{
"epoch": 3.12,
"learning_rate": 4.803700921204659e-06,
"loss": 0.0427,
"step": 24400
},
{
"epoch": 3.12,
"learning_rate": 4.791439139607124e-06,
"loss": 0.0426,
"step": 24425
},
{
"epoch": 3.12,
"learning_rate": 4.779178614364664e-06,
"loss": 0.0405,
"step": 24450
},
{
"epoch": 3.13,
"learning_rate": 4.766919419333767e-06,
"loss": 0.0411,
"step": 24475
},
{
"epoch": 3.13,
"learning_rate": 4.75466162836291e-06,
"loss": 0.0429,
"step": 24500
},
{
"epoch": 3.13,
"learning_rate": 4.742405315292113e-06,
"loss": 0.0398,
"step": 24525
},
{
"epoch": 3.14,
"learning_rate": 4.730150553952491e-06,
"loss": 0.0417,
"step": 24550
},
{
"epoch": 3.14,
"learning_rate": 4.717897418165813e-06,
"loss": 0.0413,
"step": 24575
},
{
"epoch": 3.14,
"learning_rate": 4.705645981744055e-06,
"loss": 0.0458,
"step": 24600
},
{
"epoch": 3.15,
"learning_rate": 4.693396318488958e-06,
"loss": 0.0436,
"step": 24625
},
{
"epoch": 3.15,
"learning_rate": 4.6811485021915784e-06,
"loss": 0.0446,
"step": 24650
},
{
"epoch": 3.15,
"learning_rate": 4.668902606631849e-06,
"loss": 0.0426,
"step": 24675
},
{
"epoch": 3.16,
"learning_rate": 4.6566587055781324e-06,
"loss": 0.0404,
"step": 24700
},
{
"epoch": 3.16,
"learning_rate": 4.6444168727867735e-06,
"loss": 0.0445,
"step": 24725
},
{
"epoch": 3.16,
"learning_rate": 4.6321771820016635e-06,
"loss": 0.0463,
"step": 24750
},
{
"epoch": 3.17,
"learning_rate": 4.619939706953786e-06,
"loss": 0.0425,
"step": 24775
},
{
"epoch": 3.17,
"learning_rate": 4.6077045213607765e-06,
"loss": 0.0417,
"step": 24800
},
{
"epoch": 3.17,
"learning_rate": 4.595471698926483e-06,
"loss": 0.0438,
"step": 24825
},
{
"epoch": 3.18,
"learning_rate": 4.583241313340512e-06,
"loss": 0.0403,
"step": 24850
},
{
"epoch": 3.18,
"learning_rate": 4.571013438277801e-06,
"loss": 0.0423,
"step": 24875
},
{
"epoch": 3.18,
"learning_rate": 4.5587881473981535e-06,
"loss": 0.0429,
"step": 24900
},
{
"epoch": 3.19,
"learning_rate": 4.546565514345809e-06,
"loss": 0.0433,
"step": 24925
},
{
"epoch": 3.19,
"learning_rate": 4.534345612749002e-06,
"loss": 0.0483,
"step": 24950
},
{
"epoch": 3.19,
"learning_rate": 4.5221285162195035e-06,
"loss": 0.0393,
"step": 24975
},
{
"epoch": 3.19,
"learning_rate": 4.509914298352197e-06,
"loss": 0.0406,
"step": 25000
},
{
"epoch": 3.19,
"eval_cer": 5.443632587139145,
"eval_loss": 0.061666421592235565,
"eval_runtime": 3730.868,
"eval_samples_per_second": 2.437,
"eval_steps_per_second": 0.076,
"eval_wer": 9.653842723732433,
"step": 25000
},
{
"epoch": 3.2,
"learning_rate": 4.497703032724619e-06,
"loss": 0.0439,
"step": 25025
},
{
"epoch": 3.2,
"learning_rate": 4.485494792896519e-06,
"loss": 0.0436,
"step": 25050
},
{
"epoch": 3.2,
"learning_rate": 4.473289652409431e-06,
"loss": 0.0393,
"step": 25075
},
{
"epoch": 3.21,
"learning_rate": 4.4610876847862034e-06,
"loss": 0.0402,
"step": 25100
},
{
"epoch": 3.21,
"learning_rate": 4.448888963530585e-06,
"loss": 0.0439,
"step": 25125
},
{
"epoch": 3.21,
"learning_rate": 4.436693562126762e-06,
"loss": 0.0406,
"step": 25150
},
{
"epoch": 3.22,
"learning_rate": 4.42450155403892e-06,
"loss": 0.0405,
"step": 25175
},
{
"epoch": 3.22,
"learning_rate": 4.4123130127108125e-06,
"loss": 0.0463,
"step": 25200
},
{
"epoch": 3.22,
"learning_rate": 4.4001280115652984e-06,
"loss": 0.0432,
"step": 25225
},
{
"epoch": 3.23,
"learning_rate": 4.38794662400392e-06,
"loss": 0.0405,
"step": 25250
},
{
"epoch": 3.23,
"learning_rate": 4.375768923406445e-06,
"loss": 0.0437,
"step": 25275
},
{
"epoch": 3.23,
"learning_rate": 4.363594983130435e-06,
"loss": 0.0409,
"step": 25300
},
{
"epoch": 3.24,
"learning_rate": 4.351424876510798e-06,
"loss": 0.0428,
"step": 25325
},
{
"epoch": 3.24,
"learning_rate": 4.339258676859349e-06,
"loss": 0.0438,
"step": 25350
},
{
"epoch": 3.24,
"learning_rate": 4.3270964574643695e-06,
"loss": 0.0438,
"step": 25375
},
{
"epoch": 3.25,
"learning_rate": 4.314938291590161e-06,
"loss": 0.0435,
"step": 25400
},
{
"epoch": 3.25,
"learning_rate": 4.3027842524766066e-06,
"loss": 0.043,
"step": 25425
},
{
"epoch": 3.25,
"learning_rate": 4.290634413338735e-06,
"loss": 0.0394,
"step": 25450
},
{
"epoch": 3.26,
"learning_rate": 4.278488847366268e-06,
"loss": 0.0401,
"step": 25475
},
{
"epoch": 3.26,
"learning_rate": 4.266347627723192e-06,
"loss": 0.0449,
"step": 25500
},
{
"epoch": 3.26,
"learning_rate": 4.254210827547309e-06,
"loss": 0.042,
"step": 25525
},
{
"epoch": 3.27,
"learning_rate": 4.242078519949795e-06,
"loss": 0.0428,
"step": 25550
},
{
"epoch": 3.27,
"learning_rate": 4.229950778014771e-06,
"loss": 0.0413,
"step": 25575
},
{
"epoch": 3.27,
"learning_rate": 4.217827674798845e-06,
"loss": 0.0469,
"step": 25600
},
{
"epoch": 3.27,
"learning_rate": 4.205709283330694e-06,
"loss": 0.0438,
"step": 25625
},
{
"epoch": 3.28,
"learning_rate": 4.193595676610599e-06,
"loss": 0.0405,
"step": 25650
},
{
"epoch": 3.28,
"learning_rate": 4.181486927610023e-06,
"loss": 0.0412,
"step": 25675
},
{
"epoch": 3.28,
"learning_rate": 4.169383109271174e-06,
"loss": 0.0431,
"step": 25700
},
{
"epoch": 3.29,
"learning_rate": 4.157284294506543e-06,
"loss": 0.0441,
"step": 25725
},
{
"epoch": 3.29,
"learning_rate": 4.145190556198494e-06,
"loss": 0.043,
"step": 25750
},
{
"epoch": 3.29,
"learning_rate": 4.133101967198804e-06,
"loss": 0.0403,
"step": 25775
},
{
"epoch": 3.3,
"learning_rate": 4.1210186003282275e-06,
"loss": 0.043,
"step": 25800
},
{
"epoch": 3.3,
"learning_rate": 4.108940528376071e-06,
"loss": 0.0441,
"step": 25825
},
{
"epoch": 3.3,
"learning_rate": 4.096867824099736e-06,
"loss": 0.0432,
"step": 25850
},
{
"epoch": 3.31,
"learning_rate": 4.084800560224296e-06,
"loss": 0.0427,
"step": 25875
},
{
"epoch": 3.31,
"learning_rate": 4.072738809442046e-06,
"loss": 0.0432,
"step": 25900
},
{
"epoch": 3.31,
"learning_rate": 4.060682644412075e-06,
"loss": 0.0419,
"step": 25925
},
{
"epoch": 3.32,
"learning_rate": 4.048632137759821e-06,
"loss": 0.0423,
"step": 25950
},
{
"epoch": 3.32,
"learning_rate": 4.036587362076639e-06,
"loss": 0.0471,
"step": 25975
},
{
"epoch": 3.32,
"learning_rate": 4.02454838991936e-06,
"loss": 0.041,
"step": 26000
},
{
"epoch": 3.32,
"eval_cer": 6.086712273582361,
"eval_loss": 0.061140306293964386,
"eval_runtime": 3838.2645,
"eval_samples_per_second": 2.369,
"eval_steps_per_second": 0.074,
"eval_wer": 10.383444245788228,
"step": 26000
},
{
"epoch": 3.33,
"learning_rate": 4.012515293809853e-06,
"loss": 0.0434,
"step": 26025
},
{
"epoch": 3.33,
"learning_rate": 4.000488146234592e-06,
"loss": 0.0393,
"step": 26050
},
{
"epoch": 3.33,
"learning_rate": 3.9884670196442184e-06,
"loss": 0.0435,
"step": 26075
},
{
"epoch": 3.34,
"learning_rate": 3.9764519864531026e-06,
"loss": 0.0432,
"step": 26100
},
{
"epoch": 3.34,
"learning_rate": 3.964443119038908e-06,
"loss": 0.0442,
"step": 26125
},
{
"epoch": 3.34,
"learning_rate": 3.952440489742158e-06,
"loss": 0.0427,
"step": 26150
},
{
"epoch": 3.34,
"learning_rate": 3.940444170865793e-06,
"loss": 0.0375,
"step": 26175
},
{
"epoch": 3.35,
"learning_rate": 3.928454234674748e-06,
"loss": 0.041,
"step": 26200
},
{
"epoch": 3.35,
"learning_rate": 3.916470753395499e-06,
"loss": 0.0415,
"step": 26225
},
{
"epoch": 3.35,
"learning_rate": 3.904493799215652e-06,
"loss": 0.0418,
"step": 26250
},
{
"epoch": 3.36,
"learning_rate": 3.89252344428348e-06,
"loss": 0.0367,
"step": 26275
},
{
"epoch": 3.36,
"learning_rate": 3.880559760707508e-06,
"loss": 0.0432,
"step": 26300
},
{
"epoch": 3.36,
"learning_rate": 3.868602820556079e-06,
"loss": 0.0404,
"step": 26325
},
{
"epoch": 3.37,
"learning_rate": 3.8566526958569025e-06,
"loss": 0.0372,
"step": 26350
},
{
"epoch": 3.37,
"learning_rate": 3.844709458596645e-06,
"loss": 0.0426,
"step": 26375
},
{
"epoch": 3.37,
"learning_rate": 3.832773180720475e-06,
"loss": 0.0419,
"step": 26400
},
{
"epoch": 3.38,
"learning_rate": 3.820843934131637e-06,
"loss": 0.0427,
"step": 26425
},
{
"epoch": 3.38,
"learning_rate": 3.8089217906910274e-06,
"loss": 0.044,
"step": 26450
},
{
"epoch": 3.38,
"learning_rate": 3.797006822216748e-06,
"loss": 0.0362,
"step": 26475
},
{
"epoch": 3.39,
"learning_rate": 3.7850991004836813e-06,
"loss": 0.044,
"step": 26500
},
{
"epoch": 3.39,
"learning_rate": 3.773198697223054e-06,
"loss": 0.0404,
"step": 26525
},
{
"epoch": 3.39,
"learning_rate": 3.761305684122008e-06,
"loss": 0.0431,
"step": 26550
},
{
"epoch": 3.4,
"learning_rate": 3.7494201328231693e-06,
"loss": 0.0388,
"step": 26575
},
{
"epoch": 3.4,
"learning_rate": 3.7375421149242102e-06,
"loss": 0.0446,
"step": 26600
},
{
"epoch": 3.4,
"learning_rate": 3.725671701977428e-06,
"loss": 0.0388,
"step": 26625
},
{
"epoch": 3.41,
"learning_rate": 3.7138089654893027e-06,
"loss": 0.0399,
"step": 26650
},
{
"epoch": 3.41,
"learning_rate": 3.7019539769200754e-06,
"loss": 0.0437,
"step": 26675
},
{
"epoch": 3.41,
"learning_rate": 3.6901068076833136e-06,
"loss": 0.0381,
"step": 26700
},
{
"epoch": 3.42,
"learning_rate": 3.6782675291454807e-06,
"loss": 0.0382,
"step": 26725
},
{
"epoch": 3.42,
"learning_rate": 3.6664362126255087e-06,
"loss": 0.0395,
"step": 26750
},
{
"epoch": 3.42,
"learning_rate": 3.6546129293943666e-06,
"loss": 0.0398,
"step": 26775
},
{
"epoch": 3.42,
"learning_rate": 3.6427977506746293e-06,
"loss": 0.0396,
"step": 26800
},
{
"epoch": 3.43,
"learning_rate": 3.6309907476400554e-06,
"loss": 0.0381,
"step": 26825
},
{
"epoch": 3.43,
"learning_rate": 3.619191991415146e-06,
"loss": 0.0366,
"step": 26850
},
{
"epoch": 3.43,
"learning_rate": 3.6074015530747354e-06,
"loss": 0.0391,
"step": 26875
},
{
"epoch": 3.44,
"learning_rate": 3.595619503643541e-06,
"loss": 0.0408,
"step": 26900
},
{
"epoch": 3.44,
"learning_rate": 3.583845914095749e-06,
"loss": 0.0378,
"step": 26925
},
{
"epoch": 3.44,
"learning_rate": 3.5720808553545894e-06,
"loss": 0.0412,
"step": 26950
},
{
"epoch": 3.45,
"learning_rate": 3.5603243982918946e-06,
"loss": 0.0389,
"step": 26975
},
{
"epoch": 3.45,
"learning_rate": 3.5485766137276894e-06,
"loss": 0.0388,
"step": 27000
},
{
"epoch": 3.45,
"eval_cer": 6.16414054617423,
"eval_loss": 0.06143974885344505,
"eval_runtime": 3828.3886,
"eval_samples_per_second": 2.375,
"eval_steps_per_second": 0.074,
"eval_wer": 10.388965095489466,
"step": 27000
},
{
"epoch": 3.45,
"learning_rate": 3.5368375724297495e-06,
"loss": 0.0425,
"step": 27025
},
{
"epoch": 3.46,
"learning_rate": 3.5251073451131824e-06,
"loss": 0.0395,
"step": 27050
},
{
"epoch": 3.46,
"learning_rate": 3.513386002440009e-06,
"loss": 0.0424,
"step": 27075
},
{
"epoch": 3.46,
"learning_rate": 3.501673615018717e-06,
"loss": 0.0382,
"step": 27100
},
{
"epoch": 3.47,
"learning_rate": 3.48997025340386e-06,
"loss": 0.0392,
"step": 27125
},
{
"epoch": 3.47,
"learning_rate": 3.478275988095615e-06,
"loss": 0.0411,
"step": 27150
},
{
"epoch": 3.47,
"learning_rate": 3.466590889539361e-06,
"loss": 0.042,
"step": 27175
},
{
"epoch": 3.48,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.039,
"step": 27200
},
{
"epoch": 3.48,
"learning_rate": 3.4432484741878404e-06,
"loss": 0.0376,
"step": 27225
},
{
"epoch": 3.48,
"learning_rate": 3.4315912980055433e-06,
"loss": 0.0394,
"step": 27250
},
{
"epoch": 3.49,
"learning_rate": 3.419943569800332e-06,
"loss": 0.0405,
"step": 27275
},
{
"epoch": 3.49,
"learning_rate": 3.4083053597372517e-06,
"loss": 0.0396,
"step": 27300
},
{
"epoch": 3.49,
"learning_rate": 3.396676737924014e-06,
"loss": 0.0405,
"step": 27325
},
{
"epoch": 3.5,
"learning_rate": 3.3850577744105682e-06,
"loss": 0.0408,
"step": 27350
},
{
"epoch": 3.5,
"learning_rate": 3.373448539188686e-06,
"loss": 0.038,
"step": 27375
},
{
"epoch": 3.5,
"learning_rate": 3.3618491021915334e-06,
"loss": 0.0406,
"step": 27400
},
{
"epoch": 3.5,
"learning_rate": 3.350259533293254e-06,
"loss": 0.0401,
"step": 27425
},
{
"epoch": 3.51,
"learning_rate": 3.338679902308547e-06,
"loss": 0.0405,
"step": 27450
},
{
"epoch": 3.51,
"learning_rate": 3.327110278992246e-06,
"loss": 0.0365,
"step": 27475
},
{
"epoch": 3.51,
"learning_rate": 3.3155507330389004e-06,
"loss": 0.0417,
"step": 27500
},
{
"epoch": 3.52,
"learning_rate": 3.304001334082353e-06,
"loss": 0.0391,
"step": 27525
},
{
"epoch": 3.52,
"learning_rate": 3.2924621516953195e-06,
"loss": 0.039,
"step": 27550
},
{
"epoch": 3.52,
"learning_rate": 3.2809332553889773e-06,
"loss": 0.041,
"step": 27575
},
{
"epoch": 3.53,
"learning_rate": 3.269414714612534e-06,
"loss": 0.0358,
"step": 27600
},
{
"epoch": 3.53,
"learning_rate": 3.257906598752828e-06,
"loss": 0.0391,
"step": 27625
},
{
"epoch": 3.53,
"learning_rate": 3.2464089771338856e-06,
"loss": 0.0363,
"step": 27650
},
{
"epoch": 3.54,
"learning_rate": 3.2349219190165203e-06,
"loss": 0.0399,
"step": 27675
},
{
"epoch": 3.54,
"learning_rate": 3.223445493597921e-06,
"loss": 0.0389,
"step": 27700
},
{
"epoch": 3.54,
"learning_rate": 3.211979770011211e-06,
"loss": 0.0376,
"step": 27725
},
{
"epoch": 3.55,
"learning_rate": 3.2005248173250593e-06,
"loss": 0.0359,
"step": 27750
},
{
"epoch": 3.55,
"learning_rate": 3.189080704543247e-06,
"loss": 0.0378,
"step": 27775
},
{
"epoch": 3.55,
"learning_rate": 3.177647500604252e-06,
"loss": 0.0398,
"step": 27800
},
{
"epoch": 3.56,
"learning_rate": 3.166225274380846e-06,
"loss": 0.0373,
"step": 27825
},
{
"epoch": 3.56,
"learning_rate": 3.154814094679668e-06,
"loss": 0.0398,
"step": 27850
},
{
"epoch": 3.56,
"learning_rate": 3.143414030240813e-06,
"loss": 0.04,
"step": 27875
},
{
"epoch": 3.57,
"learning_rate": 3.1320251497374187e-06,
"loss": 0.0412,
"step": 27900
},
{
"epoch": 3.57,
"learning_rate": 3.120647521775253e-06,
"loss": 0.0373,
"step": 27925
},
{
"epoch": 3.57,
"learning_rate": 3.109281214892298e-06,
"loss": 0.0421,
"step": 27950
},
{
"epoch": 3.58,
"learning_rate": 3.097926297558339e-06,
"loss": 0.0408,
"step": 27975
},
{
"epoch": 3.58,
"learning_rate": 3.0865828381745515e-06,
"loss": 0.0383,
"step": 28000
},
{
"epoch": 3.58,
"eval_cer": 6.146030830866347,
"eval_loss": 0.06108880788087845,
"eval_runtime": 3825.6759,
"eval_samples_per_second": 2.377,
"eval_steps_per_second": 0.074,
"eval_wer": 10.353716593550798,
"step": 28000
},
{
"epoch": 3.58,
"learning_rate": 3.07525090507309e-06,
"loss": 0.0417,
"step": 28025
},
{
"epoch": 3.58,
"learning_rate": 3.0639305665166724e-06,
"loss": 0.0408,
"step": 28050
},
{
"epoch": 3.59,
"learning_rate": 3.0526218906981764e-06,
"loss": 0.0407,
"step": 28075
},
{
"epoch": 3.59,
"learning_rate": 3.0413249457402206e-06,
"loss": 0.0367,
"step": 28100
},
{
"epoch": 3.59,
"learning_rate": 3.0300397996947604e-06,
"loss": 0.0381,
"step": 28125
},
{
"epoch": 3.6,
"learning_rate": 3.018766520542673e-06,
"loss": 0.0385,
"step": 28150
},
{
"epoch": 3.6,
"learning_rate": 3.0075051761933515e-06,
"loss": 0.0374,
"step": 28175
},
{
"epoch": 3.6,
"learning_rate": 2.9962558344842963e-06,
"loss": 0.0373,
"step": 28200
},
{
"epoch": 3.61,
"learning_rate": 2.9850185631806996e-06,
"loss": 0.0394,
"step": 28225
},
{
"epoch": 3.61,
"learning_rate": 2.9737934299750514e-06,
"loss": 0.0424,
"step": 28250
},
{
"epoch": 3.61,
"learning_rate": 2.9625805024867136e-06,
"loss": 0.0407,
"step": 28275
},
{
"epoch": 3.62,
"learning_rate": 2.951379848261523e-06,
"loss": 0.0391,
"step": 28300
},
{
"epoch": 3.62,
"learning_rate": 2.9401915347713927e-06,
"loss": 0.0388,
"step": 28325
},
{
"epoch": 3.62,
"learning_rate": 2.9290156294138807e-06,
"loss": 0.0401,
"step": 28350
},
{
"epoch": 3.63,
"learning_rate": 2.9178521995118154e-06,
"loss": 0.039,
"step": 28375
},
{
"epoch": 3.63,
"learning_rate": 2.906701312312861e-06,
"loss": 0.0418,
"step": 28400
},
{
"epoch": 3.63,
"learning_rate": 2.8955630349891306e-06,
"loss": 0.0433,
"step": 28425
},
{
"epoch": 3.64,
"learning_rate": 2.88443743463678e-06,
"loss": 0.0375,
"step": 28450
},
{
"epoch": 3.64,
"learning_rate": 2.8733245782755937e-06,
"loss": 0.0402,
"step": 28475
},
{
"epoch": 3.64,
"learning_rate": 2.862224532848591e-06,
"loss": 0.0365,
"step": 28500
},
{
"epoch": 3.65,
"learning_rate": 2.851137365221617e-06,
"loss": 0.0413,
"step": 28525
},
{
"epoch": 3.65,
"learning_rate": 2.840063142182941e-06,
"loss": 0.0433,
"step": 28550
},
{
"epoch": 3.65,
"learning_rate": 2.829001930442862e-06,
"loss": 0.0397,
"step": 28575
},
{
"epoch": 3.65,
"learning_rate": 2.817953796633289e-06,
"loss": 0.0372,
"step": 28600
},
{
"epoch": 3.66,
"learning_rate": 2.806918807307363e-06,
"loss": 0.0403,
"step": 28625
},
{
"epoch": 3.66,
"learning_rate": 2.7958970289390317e-06,
"loss": 0.0405,
"step": 28650
},
{
"epoch": 3.66,
"learning_rate": 2.784888527922667e-06,
"loss": 0.0364,
"step": 28675
},
{
"epoch": 3.67,
"learning_rate": 2.77389337057266e-06,
"loss": 0.0389,
"step": 28700
},
{
"epoch": 3.67,
"learning_rate": 2.7629116231230156e-06,
"loss": 0.0364,
"step": 28725
},
{
"epoch": 3.67,
"learning_rate": 2.7519433517269665e-06,
"loss": 0.0409,
"step": 28750
},
{
"epoch": 3.68,
"learning_rate": 2.740988622456562e-06,
"loss": 0.0441,
"step": 28775
},
{
"epoch": 3.68,
"learning_rate": 2.7300475013022666e-06,
"loss": 0.0373,
"step": 28800
},
{
"epoch": 3.68,
"learning_rate": 2.719120054172586e-06,
"loss": 0.0393,
"step": 28825
},
{
"epoch": 3.69,
"learning_rate": 2.7082063468936427e-06,
"loss": 0.0374,
"step": 28850
},
{
"epoch": 3.69,
"learning_rate": 2.6973064452088e-06,
"loss": 0.0392,
"step": 28875
},
{
"epoch": 3.69,
"learning_rate": 2.68642041477825e-06,
"loss": 0.0395,
"step": 28900
},
{
"epoch": 3.7,
"learning_rate": 2.6755483211786294e-06,
"loss": 0.0374,
"step": 28925
},
{
"epoch": 3.7,
"learning_rate": 2.6646902299026183e-06,
"loss": 0.0394,
"step": 28950
},
{
"epoch": 3.7,
"learning_rate": 2.6538462063585478e-06,
"loss": 0.0389,
"step": 28975
},
{
"epoch": 3.71,
"learning_rate": 2.6430163158700116e-06,
"loss": 0.0401,
"step": 29000
},
{
"epoch": 3.71,
"eval_cer": 6.957641745480887,
"eval_loss": 0.06025030463933945,
"eval_runtime": 3930.21,
"eval_samples_per_second": 2.314,
"eval_steps_per_second": 0.073,
"eval_wer": 11.069728331726624,
"step": 29000
},
{
"epoch": 3.71,
"learning_rate": 2.6322006236754603e-06,
"loss": 0.0399,
"step": 29025
},
{
"epoch": 3.71,
"learning_rate": 2.621399194927817e-06,
"loss": 0.0387,
"step": 29050
},
{
"epoch": 3.72,
"learning_rate": 2.6106120946940826e-06,
"loss": 0.0385,
"step": 29075
},
{
"epoch": 3.72,
"learning_rate": 2.5998393879549444e-06,
"loss": 0.0378,
"step": 29100
},
{
"epoch": 3.72,
"learning_rate": 2.589081139604387e-06,
"loss": 0.0419,
"step": 29125
},
{
"epoch": 3.73,
"learning_rate": 2.5783374144492946e-06,
"loss": 0.0378,
"step": 29150
},
{
"epoch": 3.73,
"learning_rate": 2.5676082772090673e-06,
"loss": 0.0364,
"step": 29175
},
{
"epoch": 3.73,
"learning_rate": 2.5568937925152272e-06,
"loss": 0.0378,
"step": 29200
},
{
"epoch": 3.73,
"learning_rate": 2.5461940249110295e-06,
"loss": 0.0394,
"step": 29225
},
{
"epoch": 3.74,
"learning_rate": 2.5355090388510806e-06,
"loss": 0.0398,
"step": 29250
},
{
"epoch": 3.74,
"learning_rate": 2.5248388987009374e-06,
"loss": 0.0394,
"step": 29275
},
{
"epoch": 3.74,
"learning_rate": 2.5141836687367273e-06,
"loss": 0.0421,
"step": 29300
},
{
"epoch": 3.75,
"learning_rate": 2.5035434131447673e-06,
"loss": 0.0396,
"step": 29325
},
{
"epoch": 3.75,
"learning_rate": 2.4929181960211553e-06,
"loss": 0.0383,
"step": 29350
},
{
"epoch": 3.75,
"learning_rate": 2.482308081371413e-06,
"loss": 0.0362,
"step": 29375
},
{
"epoch": 3.76,
"learning_rate": 2.471713133110078e-06,
"loss": 0.035,
"step": 29400
},
{
"epoch": 3.76,
"learning_rate": 2.461133415060327e-06,
"loss": 0.0377,
"step": 29425
},
{
"epoch": 3.76,
"learning_rate": 2.4505689909535967e-06,
"loss": 0.04,
"step": 29450
},
{
"epoch": 3.77,
"learning_rate": 2.4400199244291884e-06,
"loss": 0.0408,
"step": 29475
},
{
"epoch": 3.77,
"learning_rate": 2.4294862790338924e-06,
"loss": 0.0372,
"step": 29500
},
{
"epoch": 3.77,
"learning_rate": 2.4189681182216023e-06,
"loss": 0.0376,
"step": 29525
},
{
"epoch": 3.78,
"learning_rate": 2.4084655053529337e-06,
"loss": 0.0388,
"step": 29550
},
{
"epoch": 3.78,
"learning_rate": 2.3979785036948468e-06,
"loss": 0.036,
"step": 29575
},
{
"epoch": 3.78,
"learning_rate": 2.387507176420256e-06,
"loss": 0.0409,
"step": 29600
},
{
"epoch": 3.79,
"learning_rate": 2.3770515866076564e-06,
"loss": 0.0383,
"step": 29625
},
{
"epoch": 3.79,
"learning_rate": 2.366611797240741e-06,
"loss": 0.0419,
"step": 29650
},
{
"epoch": 3.79,
"learning_rate": 2.3561878712080213e-06,
"loss": 0.0388,
"step": 29675
},
{
"epoch": 3.8,
"learning_rate": 2.345779871302453e-06,
"loss": 0.0355,
"step": 29700
},
{
"epoch": 3.8,
"learning_rate": 2.335387860221051e-06,
"loss": 0.037,
"step": 29725
},
{
"epoch": 3.8,
"learning_rate": 2.325011900564515e-06,
"loss": 0.0404,
"step": 29750
},
{
"epoch": 3.81,
"learning_rate": 2.3146520548368524e-06,
"loss": 0.0348,
"step": 29775
},
{
"epoch": 3.81,
"learning_rate": 2.304308385444999e-06,
"loss": 0.0395,
"step": 29800
},
{
"epoch": 3.81,
"learning_rate": 2.293980954698452e-06,
"loss": 0.0357,
"step": 29825
},
{
"epoch": 3.81,
"learning_rate": 2.2836698248088814e-06,
"loss": 0.0368,
"step": 29850
},
{
"epoch": 3.82,
"learning_rate": 2.273375057889769e-06,
"loss": 0.0346,
"step": 29875
},
{
"epoch": 3.82,
"learning_rate": 2.263096715956019e-06,
"loss": 0.0389,
"step": 29900
},
{
"epoch": 3.82,
"learning_rate": 2.2528348609235963e-06,
"loss": 0.0378,
"step": 29925
},
{
"epoch": 3.83,
"learning_rate": 2.2425895546091534e-06,
"loss": 0.0397,
"step": 29950
},
{
"epoch": 3.83,
"learning_rate": 2.232360858729649e-06,
"loss": 0.0369,
"step": 29975
},
{
"epoch": 3.83,
"learning_rate": 2.2221488349019903e-06,
"loss": 0.0343,
"step": 30000
},
{
"epoch": 3.83,
"eval_cer": 7.1917744933899534,
"eval_loss": 0.06130935996770859,
"eval_runtime": 3953.1262,
"eval_samples_per_second": 2.3,
"eval_steps_per_second": 0.072,
"eval_wer": 11.224312123361264,
"step": 30000
},
{
"epoch": 3.84,
"learning_rate": 2.211953544642641e-06,
"loss": 0.036,
"step": 30025
},
{
"epoch": 3.84,
"learning_rate": 2.2017750493672704e-06,
"loss": 0.0365,
"step": 30050
},
{
"epoch": 3.84,
"learning_rate": 2.191613410390378e-06,
"loss": 0.0405,
"step": 30075
},
{
"epoch": 3.85,
"learning_rate": 2.181468688924916e-06,
"loss": 0.0372,
"step": 30100
},
{
"epoch": 3.85,
"learning_rate": 2.171340946081934e-06,
"loss": 0.0371,
"step": 30125
},
{
"epoch": 3.85,
"learning_rate": 2.1612302428701993e-06,
"loss": 0.038,
"step": 30150
},
{
"epoch": 3.86,
"learning_rate": 2.1511366401958293e-06,
"loss": 0.038,
"step": 30175
},
{
"epoch": 3.86,
"learning_rate": 2.1410601988619394e-06,
"loss": 0.0414,
"step": 30200
},
{
"epoch": 3.86,
"learning_rate": 2.1310009795682573e-06,
"loss": 0.0356,
"step": 30225
},
{
"epoch": 3.87,
"learning_rate": 2.1209590429107734e-06,
"loss": 0.0413,
"step": 30250
},
{
"epoch": 3.87,
"learning_rate": 2.1109344493813638e-06,
"loss": 0.0396,
"step": 30275
},
{
"epoch": 3.87,
"learning_rate": 2.1009272593674323e-06,
"loss": 0.0374,
"step": 30300
},
{
"epoch": 3.88,
"learning_rate": 2.090937533151546e-06,
"loss": 0.0415,
"step": 30325
},
{
"epoch": 3.88,
"learning_rate": 2.0809653309110685e-06,
"loss": 0.0383,
"step": 30350
},
{
"epoch": 3.88,
"learning_rate": 2.071010712717806e-06,
"loss": 0.0413,
"step": 30375
},
{
"epoch": 3.88,
"learning_rate": 2.061073738537635e-06,
"loss": 0.0378,
"step": 30400
},
{
"epoch": 3.89,
"learning_rate": 2.0511544682301473e-06,
"loss": 0.0358,
"step": 30425
},
{
"epoch": 3.89,
"learning_rate": 2.0412529615482867e-06,
"loss": 0.0414,
"step": 30450
},
{
"epoch": 3.89,
"learning_rate": 2.03136927813799e-06,
"loss": 0.0411,
"step": 30475
},
{
"epoch": 3.9,
"learning_rate": 2.0215034775378336e-06,
"loss": 0.0391,
"step": 30500
},
{
"epoch": 3.9,
"learning_rate": 2.011655619178662e-06,
"loss": 0.0343,
"step": 30525
},
{
"epoch": 3.9,
"learning_rate": 2.0018257623832393e-06,
"loss": 0.0335,
"step": 30550
},
{
"epoch": 3.91,
"learning_rate": 1.992013966365891e-06,
"loss": 0.0397,
"step": 30575
},
{
"epoch": 3.91,
"learning_rate": 1.982220290232143e-06,
"loss": 0.0352,
"step": 30600
},
{
"epoch": 3.91,
"learning_rate": 1.972444792978373e-06,
"loss": 0.0388,
"step": 30625
},
{
"epoch": 3.92,
"learning_rate": 1.962687533491446e-06,
"loss": 0.0381,
"step": 30650
},
{
"epoch": 3.92,
"learning_rate": 1.9529485705483642e-06,
"loss": 0.0381,
"step": 30675
},
{
"epoch": 3.92,
"learning_rate": 1.9432279628159188e-06,
"loss": 0.0386,
"step": 30700
},
{
"epoch": 3.93,
"learning_rate": 1.933525768850321e-06,
"loss": 0.0406,
"step": 30725
},
{
"epoch": 3.93,
"learning_rate": 1.9238420470968665e-06,
"loss": 0.0389,
"step": 30750
},
{
"epoch": 3.93,
"learning_rate": 1.914176855889573e-06,
"loss": 0.0387,
"step": 30775
},
{
"epoch": 3.94,
"learning_rate": 1.9045302534508298e-06,
"loss": 0.036,
"step": 30800
},
{
"epoch": 3.94,
"learning_rate": 1.8949022978910541e-06,
"loss": 0.0376,
"step": 30825
},
{
"epoch": 3.94,
"learning_rate": 1.8852930472083304e-06,
"loss": 0.0398,
"step": 30850
},
{
"epoch": 3.95,
"learning_rate": 1.875702559288069e-06,
"loss": 0.0387,
"step": 30875
},
{
"epoch": 3.95,
"learning_rate": 1.8661308919026533e-06,
"loss": 0.0356,
"step": 30900
},
{
"epoch": 3.95,
"learning_rate": 1.856578102711093e-06,
"loss": 0.0351,
"step": 30925
},
{
"epoch": 3.96,
"learning_rate": 1.847044249258681e-06,
"loss": 0.0378,
"step": 30950
},
{
"epoch": 3.96,
"learning_rate": 1.8375293889766382e-06,
"loss": 0.0418,
"step": 30975
},
{
"epoch": 3.96,
"learning_rate": 1.8280335791817733e-06,
"loss": 0.0357,
"step": 31000
},
{
"epoch": 3.96,
"eval_cer": 7.312813917131421,
"eval_loss": 0.06026212498545647,
"eval_runtime": 3969.9658,
"eval_samples_per_second": 2.29,
"eval_steps_per_second": 0.072,
"eval_wer": 11.331331671416013,
"step": 31000
},
{
"epoch": 3.96,
"learning_rate": 1.8185568770761368e-06,
"loss": 0.0389,
"step": 31025
},
{
"epoch": 3.97,
"learning_rate": 1.809099339746674e-06,
"loss": 0.0357,
"step": 31050
},
{
"epoch": 3.97,
"learning_rate": 1.799661024164887e-06,
"loss": 0.0366,
"step": 31075
},
{
"epoch": 3.97,
"learning_rate": 1.790241987186485e-06,
"loss": 0.0366,
"step": 31100
},
{
"epoch": 3.98,
"learning_rate": 1.7808422855510438e-06,
"loss": 0.0385,
"step": 31125
},
{
"epoch": 3.98,
"learning_rate": 1.7714619758816653e-06,
"loss": 0.0375,
"step": 31150
},
{
"epoch": 3.98,
"learning_rate": 1.7621011146846335e-06,
"loss": 0.0369,
"step": 31175
},
{
"epoch": 3.99,
"learning_rate": 1.7527597583490825e-06,
"loss": 0.0367,
"step": 31200
},
{
"epoch": 3.99,
"learning_rate": 1.743437963146642e-06,
"loss": 0.0378,
"step": 31225
},
{
"epoch": 3.99,
"learning_rate": 1.7341357852311175e-06,
"loss": 0.0336,
"step": 31250
},
{
"epoch": 4.0,
"learning_rate": 1.7248532806381291e-06,
"loss": 0.0367,
"step": 31275
},
{
"epoch": 4.0,
"learning_rate": 1.7155905052847938e-06,
"loss": 0.0361,
"step": 31300
},
{
"epoch": 4.0,
"learning_rate": 1.7063475149693814e-06,
"loss": 0.0293,
"step": 31325
},
{
"epoch": 4.01,
"learning_rate": 1.697124365370974e-06,
"loss": 0.031,
"step": 31350
},
{
"epoch": 4.01,
"learning_rate": 1.6879211120491407e-06,
"loss": 0.029,
"step": 31375
},
{
"epoch": 4.01,
"learning_rate": 1.6787378104435931e-06,
"loss": 0.0321,
"step": 31400
},
{
"epoch": 4.02,
"learning_rate": 1.6695745158738503e-06,
"loss": 0.0297,
"step": 31425
},
{
"epoch": 4.02,
"learning_rate": 1.6604312835389202e-06,
"loss": 0.0308,
"step": 31450
},
{
"epoch": 4.02,
"learning_rate": 1.65130816851695e-06,
"loss": 0.032,
"step": 31475
},
{
"epoch": 4.03,
"learning_rate": 1.642205225764908e-06,
"loss": 0.0309,
"step": 31500
},
{
"epoch": 4.03,
"learning_rate": 1.6331225101182403e-06,
"loss": 0.0307,
"step": 31525
},
{
"epoch": 4.03,
"learning_rate": 1.6240600762905485e-06,
"loss": 0.0344,
"step": 31550
},
{
"epoch": 4.04,
"learning_rate": 1.615017978873259e-06,
"loss": 0.0321,
"step": 31575
},
{
"epoch": 4.04,
"learning_rate": 1.6059962723352906e-06,
"loss": 0.0312,
"step": 31600
},
{
"epoch": 4.04,
"learning_rate": 1.596995011022735e-06,
"loss": 0.0265,
"step": 31625
},
{
"epoch": 4.04,
"learning_rate": 1.588014249158516e-06,
"loss": 0.0303,
"step": 31650
},
{
"epoch": 4.05,
"learning_rate": 1.5790540408420746e-06,
"loss": 0.0321,
"step": 31675
},
{
"epoch": 4.05,
"learning_rate": 1.570114440049037e-06,
"loss": 0.0287,
"step": 31700
},
{
"epoch": 4.05,
"learning_rate": 1.561195500630891e-06,
"loss": 0.0294,
"step": 31725
},
{
"epoch": 4.06,
"learning_rate": 1.5522972763146653e-06,
"loss": 0.0307,
"step": 31750
},
{
"epoch": 4.06,
"learning_rate": 1.5434198207025986e-06,
"loss": 0.0286,
"step": 31775
},
{
"epoch": 4.06,
"learning_rate": 1.5345631872718214e-06,
"loss": 0.0298,
"step": 31800
},
{
"epoch": 4.07,
"learning_rate": 1.5257274293740332e-06,
"loss": 0.0283,
"step": 31825
},
{
"epoch": 4.07,
"learning_rate": 1.5169126002351791e-06,
"loss": 0.0311,
"step": 31850
},
{
"epoch": 4.07,
"learning_rate": 1.508118752955136e-06,
"loss": 0.0324,
"step": 31875
},
{
"epoch": 4.08,
"learning_rate": 1.4993459405073829e-06,
"loss": 0.0296,
"step": 31900
},
{
"epoch": 4.08,
"learning_rate": 1.4905942157386854e-06,
"loss": 0.03,
"step": 31925
},
{
"epoch": 4.08,
"learning_rate": 1.4818636313687868e-06,
"loss": 0.0306,
"step": 31950
},
{
"epoch": 4.09,
"learning_rate": 1.4731542399900696e-06,
"loss": 0.0267,
"step": 31975
},
{
"epoch": 4.09,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.0313,
"step": 32000
},
{
"epoch": 4.09,
"eval_cer": 7.387100708496414,
"eval_loss": 0.06238727271556854,
"eval_runtime": 3970.9564,
"eval_samples_per_second": 2.29,
"eval_steps_per_second": 0.072,
"eval_wer": 11.386115487682135,
"step": 32000
},
{
"epoch": 4.09,
"learning_rate": 1.4557992459371084e-06,
"loss": 0.0293,
"step": 32025
},
{
"epoch": 4.1,
"learning_rate": 1.4471537478080516e-06,
"loss": 0.0306,
"step": 32050
},
{
"epoch": 4.1,
"learning_rate": 1.4385296517599334e-06,
"loss": 0.0295,
"step": 32075
},
{
"epoch": 4.1,
"learning_rate": 1.429927009743659e-06,
"loss": 0.0301,
"step": 32100
},
{
"epoch": 4.11,
"learning_rate": 1.4213458735809072e-06,
"loss": 0.0309,
"step": 32125
},
{
"epoch": 4.11,
"learning_rate": 1.412786294963801e-06,
"loss": 0.0321,
"step": 32150
},
{
"epoch": 4.11,
"learning_rate": 1.4042483254546025e-06,
"loss": 0.0311,
"step": 32175
},
{
"epoch": 4.11,
"learning_rate": 1.395732016485406e-06,
"loss": 0.0291,
"step": 32200
},
{
"epoch": 4.12,
"learning_rate": 1.387237419357821e-06,
"loss": 0.0304,
"step": 32225
},
{
"epoch": 4.12,
"learning_rate": 1.3787645852426663e-06,
"loss": 0.0303,
"step": 32250
},
{
"epoch": 4.12,
"learning_rate": 1.3703135651796633e-06,
"loss": 0.0324,
"step": 32275
},
{
"epoch": 4.13,
"learning_rate": 1.3618844100771256e-06,
"loss": 0.0298,
"step": 32300
},
{
"epoch": 4.13,
"learning_rate": 1.3534771707116578e-06,
"loss": 0.0314,
"step": 32325
},
{
"epoch": 4.13,
"learning_rate": 1.345091897727842e-06,
"loss": 0.0309,
"step": 32350
},
{
"epoch": 4.14,
"learning_rate": 1.3367286416379366e-06,
"loss": 0.0285,
"step": 32375
},
{
"epoch": 4.14,
"learning_rate": 1.3283874528215735e-06,
"loss": 0.0296,
"step": 32400
},
{
"epoch": 4.14,
"learning_rate": 1.3200683815254506e-06,
"loss": 0.0318,
"step": 32425
},
{
"epoch": 4.15,
"learning_rate": 1.3117714778630358e-06,
"loss": 0.0338,
"step": 32450
},
{
"epoch": 4.15,
"learning_rate": 1.3034967918142566e-06,
"loss": 0.028,
"step": 32475
},
{
"epoch": 4.15,
"learning_rate": 1.2952443732252058e-06,
"loss": 0.0304,
"step": 32500
},
{
"epoch": 4.16,
"learning_rate": 1.2870142718078366e-06,
"loss": 0.0312,
"step": 32525
},
{
"epoch": 4.16,
"learning_rate": 1.2788065371396652e-06,
"loss": 0.0294,
"step": 32550
},
{
"epoch": 4.16,
"learning_rate": 1.2706212186634765e-06,
"loss": 0.0313,
"step": 32575
},
{
"epoch": 4.17,
"learning_rate": 1.2624583656870153e-06,
"loss": 0.0288,
"step": 32600
},
{
"epoch": 4.17,
"learning_rate": 1.2543180273827044e-06,
"loss": 0.0296,
"step": 32625
},
{
"epoch": 4.17,
"learning_rate": 1.2462002527873301e-06,
"loss": 0.0296,
"step": 32650
},
{
"epoch": 4.18,
"learning_rate": 1.238105090801761e-06,
"loss": 0.0274,
"step": 32675
},
{
"epoch": 4.18,
"learning_rate": 1.2300325901906529e-06,
"loss": 0.0297,
"step": 32700
},
{
"epoch": 4.18,
"learning_rate": 1.2219827995821449e-06,
"loss": 0.0288,
"step": 32725
},
{
"epoch": 4.19,
"learning_rate": 1.2139557674675767e-06,
"loss": 0.0293,
"step": 32750
},
{
"epoch": 4.19,
"learning_rate": 1.2059515422011936e-06,
"loss": 0.0281,
"step": 32775
},
{
"epoch": 4.19,
"learning_rate": 1.1979701719998454e-06,
"loss": 0.0303,
"step": 32800
},
{
"epoch": 4.19,
"learning_rate": 1.1900117049427173e-06,
"loss": 0.0309,
"step": 32825
},
{
"epoch": 4.2,
"learning_rate": 1.1820761889710175e-06,
"loss": 0.0307,
"step": 32850
},
{
"epoch": 4.2,
"learning_rate": 1.1741636718877053e-06,
"loss": 0.0291,
"step": 32875
},
{
"epoch": 4.2,
"learning_rate": 1.1662742013571926e-06,
"loss": 0.0299,
"step": 32900
},
{
"epoch": 4.21,
"learning_rate": 1.1584078249050623e-06,
"loss": 0.0285,
"step": 32925
},
{
"epoch": 4.21,
"learning_rate": 1.1505645899177786e-06,
"loss": 0.0299,
"step": 32950
},
{
"epoch": 4.21,
"learning_rate": 1.1427445436424033e-06,
"loss": 0.032,
"step": 32975
},
{
"epoch": 4.22,
"learning_rate": 1.134947733186315e-06,
"loss": 0.0281,
"step": 33000
},
{
"epoch": 4.22,
"eval_cer": 7.870519231408898,
"eval_loss": 0.0625627338886261,
"eval_runtime": 3966.1764,
"eval_samples_per_second": 2.293,
"eval_steps_per_second": 0.072,
"eval_wer": 11.82481069855736,
"step": 33000
},
{
"epoch": 4.22,
"learning_rate": 1.1271742055169176e-06,
"loss": 0.0276,
"step": 33025
},
{
"epoch": 4.22,
"learning_rate": 1.1194240074613617e-06,
"loss": 0.0305,
"step": 33050
},
{
"epoch": 4.23,
"learning_rate": 1.111697185706262e-06,
"loss": 0.0257,
"step": 33075
},
{
"epoch": 4.23,
"learning_rate": 1.1039937867974166e-06,
"loss": 0.0264,
"step": 33100
},
{
"epoch": 4.23,
"learning_rate": 1.0963138571395277e-06,
"loss": 0.0297,
"step": 33125
},
{
"epoch": 4.24,
"learning_rate": 1.0886574429959185e-06,
"loss": 0.0276,
"step": 33150
},
{
"epoch": 4.24,
"learning_rate": 1.0810245904882566e-06,
"loss": 0.028,
"step": 33175
},
{
"epoch": 4.24,
"learning_rate": 1.0734153455962765e-06,
"loss": 0.0306,
"step": 33200
},
{
"epoch": 4.25,
"learning_rate": 1.0658297541575009e-06,
"loss": 0.0267,
"step": 33225
},
{
"epoch": 4.25,
"learning_rate": 1.058267861866969e-06,
"loss": 0.0295,
"step": 33250
},
{
"epoch": 4.25,
"learning_rate": 1.0507297142769556e-06,
"loss": 0.0305,
"step": 33275
},
{
"epoch": 4.26,
"learning_rate": 1.0432153567966985e-06,
"loss": 0.0275,
"step": 33300
},
{
"epoch": 4.26,
"learning_rate": 1.0357248346921305e-06,
"loss": 0.0306,
"step": 33325
},
{
"epoch": 4.26,
"learning_rate": 1.0282581930855933e-06,
"loss": 0.0288,
"step": 33350
},
{
"epoch": 4.27,
"learning_rate": 1.0208154769555828e-06,
"loss": 0.0273,
"step": 33375
},
{
"epoch": 4.27,
"learning_rate": 1.013396731136465e-06,
"loss": 0.031,
"step": 33400
},
{
"epoch": 4.27,
"learning_rate": 1.0060020003182102e-06,
"loss": 0.0281,
"step": 33425
},
{
"epoch": 4.27,
"learning_rate": 9.986313290461287e-07,
"loss": 0.0282,
"step": 33450
},
{
"epoch": 4.28,
"learning_rate": 9.912847617205923e-07,
"loss": 0.0305,
"step": 33475
},
{
"epoch": 4.28,
"learning_rate": 9.83962342596776e-07,
"loss": 0.0272,
"step": 33500
},
{
"epoch": 4.28,
"learning_rate": 9.766641157843858e-07,
"loss": 0.0307,
"step": 33525
},
{
"epoch": 4.29,
"learning_rate": 9.693901252473953e-07,
"loss": 0.0293,
"step": 33550
},
{
"epoch": 4.29,
"learning_rate": 9.621404148037827e-07,
"loss": 0.0289,
"step": 33575
},
{
"epoch": 4.29,
"learning_rate": 9.549150281252633e-07,
"loss": 0.0269,
"step": 33600
},
{
"epoch": 4.3,
"learning_rate": 9.477140087370268e-07,
"loss": 0.0323,
"step": 33625
},
{
"epoch": 4.3,
"learning_rate": 9.405374000174772e-07,
"loss": 0.0298,
"step": 33650
},
{
"epoch": 4.3,
"learning_rate": 9.333852451979692e-07,
"loss": 0.03,
"step": 33675
},
{
"epoch": 4.31,
"learning_rate": 9.262575873625529e-07,
"loss": 0.0297,
"step": 33700
},
{
"epoch": 4.31,
"learning_rate": 9.191544694477067e-07,
"loss": 0.0321,
"step": 33725
},
{
"epoch": 4.31,
"learning_rate": 9.120759342420821e-07,
"loss": 0.0292,
"step": 33750
},
{
"epoch": 4.32,
"learning_rate": 9.05022024386249e-07,
"loss": 0.0293,
"step": 33775
},
{
"epoch": 4.32,
"learning_rate": 8.979927823724321e-07,
"loss": 0.0312,
"step": 33800
},
{
"epoch": 4.32,
"learning_rate": 8.909882505442641e-07,
"loss": 0.0302,
"step": 33825
},
{
"epoch": 4.33,
"learning_rate": 8.840084710965202e-07,
"loss": 0.0281,
"step": 33850
},
{
"epoch": 4.33,
"learning_rate": 8.770534860748747e-07,
"loss": 0.028,
"step": 33875
},
{
"epoch": 4.33,
"learning_rate": 8.701233373756352e-07,
"loss": 0.0301,
"step": 33900
},
{
"epoch": 4.34,
"learning_rate": 8.63218066745501e-07,
"loss": 0.0301,
"step": 33925
},
{
"epoch": 4.34,
"learning_rate": 8.563377157813102e-07,
"loss": 0.0282,
"step": 33950
},
{
"epoch": 4.34,
"learning_rate": 8.494823259297808e-07,
"loss": 0.0324,
"step": 33975
},
{
"epoch": 4.34,
"learning_rate": 8.426519384872733e-07,
"loss": 0.0298,
"step": 34000
},
{
"epoch": 4.34,
"eval_cer": 8.336012832026848,
"eval_loss": 0.06292907893657684,
"eval_runtime": 4025.2254,
"eval_samples_per_second": 2.259,
"eval_steps_per_second": 0.071,
"eval_wer": 12.236751022418897,
"step": 34000
},
{
"epoch": 4.35,
"learning_rate": 8.358465945995325e-07,
"loss": 0.0331,
"step": 34025
},
{
"epoch": 4.35,
"learning_rate": 8.290663352614386e-07,
"loss": 0.0278,
"step": 34050
},
{
"epoch": 4.35,
"learning_rate": 8.223112013167717e-07,
"loss": 0.0322,
"step": 34075
},
{
"epoch": 4.36,
"learning_rate": 8.155812334579532e-07,
"loss": 0.0296,
"step": 34100
},
{
"epoch": 4.36,
"learning_rate": 8.088764722258097e-07,
"loss": 0.0282,
"step": 34125
},
{
"epoch": 4.36,
"learning_rate": 8.021969580093231e-07,
"loss": 0.0308,
"step": 34150
},
{
"epoch": 4.37,
"learning_rate": 7.955427310453856e-07,
"loss": 0.0261,
"step": 34175
},
{
"epoch": 4.37,
"learning_rate": 7.88913831418568e-07,
"loss": 0.0295,
"step": 34200
},
{
"epoch": 4.37,
"learning_rate": 7.823102990608644e-07,
"loss": 0.0311,
"step": 34225
},
{
"epoch": 4.38,
"learning_rate": 7.757321737514645e-07,
"loss": 0.0306,
"step": 34250
},
{
"epoch": 4.38,
"learning_rate": 7.691794951165033e-07,
"loss": 0.0279,
"step": 34275
},
{
"epoch": 4.38,
"learning_rate": 7.626523026288279e-07,
"loss": 0.0272,
"step": 34300
},
{
"epoch": 4.39,
"learning_rate": 7.561506356077603e-07,
"loss": 0.0269,
"step": 34325
},
{
"epoch": 4.39,
"learning_rate": 7.496745332188555e-07,
"loss": 0.0293,
"step": 34350
},
{
"epoch": 4.39,
"learning_rate": 7.43224034473674e-07,
"loss": 0.0332,
"step": 34375
},
{
"epoch": 4.4,
"learning_rate": 7.367991782295392e-07,
"loss": 0.0299,
"step": 34400
},
{
"epoch": 4.4,
"learning_rate": 7.304000031893049e-07,
"loss": 0.029,
"step": 34425
},
{
"epoch": 4.4,
"learning_rate": 7.240265479011249e-07,
"loss": 0.03,
"step": 34450
},
{
"epoch": 4.41,
"learning_rate": 7.176788507582178e-07,
"loss": 0.0294,
"step": 34475
},
{
"epoch": 4.41,
"learning_rate": 7.113569499986401e-07,
"loss": 0.0308,
"step": 34500
},
{
"epoch": 4.41,
"learning_rate": 7.050608837050487e-07,
"loss": 0.0304,
"step": 34525
},
{
"epoch": 4.42,
"learning_rate": 6.987906898044783e-07,
"loss": 0.0302,
"step": 34550
},
{
"epoch": 4.42,
"learning_rate": 6.925464060681097e-07,
"loss": 0.0287,
"step": 34575
},
{
"epoch": 4.42,
"learning_rate": 6.863280701110409e-07,
"loss": 0.026,
"step": 34600
},
{
"epoch": 4.42,
"learning_rate": 6.801357193920666e-07,
"loss": 0.0291,
"step": 34625
},
{
"epoch": 4.43,
"learning_rate": 6.739693912134443e-07,
"loss": 0.0301,
"step": 34650
},
{
"epoch": 4.43,
"learning_rate": 6.678291227206752e-07,
"loss": 0.028,
"step": 34675
},
{
"epoch": 4.43,
"learning_rate": 6.617149509022807e-07,
"loss": 0.0305,
"step": 34700
},
{
"epoch": 4.44,
"learning_rate": 6.556269125895731e-07,
"loss": 0.0291,
"step": 34725
},
{
"epoch": 4.44,
"learning_rate": 6.495650444564433e-07,
"loss": 0.0306,
"step": 34750
},
{
"epoch": 4.44,
"learning_rate": 6.435293830191325e-07,
"loss": 0.029,
"step": 34775
},
{
"epoch": 4.45,
"learning_rate": 6.375199646360142e-07,
"loss": 0.0293,
"step": 34800
},
{
"epoch": 4.45,
"learning_rate": 6.315368255073784e-07,
"loss": 0.0273,
"step": 34825
},
{
"epoch": 4.45,
"learning_rate": 6.255800016752089e-07,
"loss": 0.0295,
"step": 34850
},
{
"epoch": 4.46,
"learning_rate": 6.196495290229676e-07,
"loss": 0.0276,
"step": 34875
},
{
"epoch": 4.46,
"learning_rate": 6.137454432753798e-07,
"loss": 0.0314,
"step": 34900
},
{
"epoch": 4.46,
"learning_rate": 6.078677799982163e-07,
"loss": 0.0306,
"step": 34925
},
{
"epoch": 4.47,
"learning_rate": 6.020165745980855e-07,
"loss": 0.0295,
"step": 34950
},
{
"epoch": 4.47,
"learning_rate": 5.961918623222113e-07,
"loss": 0.0267,
"step": 34975
},
{
"epoch": 4.47,
"learning_rate": 5.903936782582253e-07,
"loss": 0.0282,
"step": 35000
},
{
"epoch": 4.47,
"eval_cer": 8.783951096376947,
"eval_loss": 0.06273001432418823,
"eval_runtime": 4031.785,
"eval_samples_per_second": 2.255,
"eval_steps_per_second": 0.071,
"eval_wer": 12.627032628221734,
"step": 35000
},
{
"epoch": 4.47,
"step": 35000,
"total_flos": 6.464312967168e+20,
"train_loss": 0.12201568590232305,
"train_runtime": 634981.2056,
"train_samples_per_second": 4.032,
"train_steps_per_second": 0.063
}
],
"max_steps": 40000,
"num_train_epochs": 6,
"total_flos": 6.464312967168e+20,
"trial_name": null,
"trial_params": null
}