whisper-largev2-mls-it / trainer_state.json
sgangireddy's picture
End of training
80b1b7c
{
"best_metric": 8.335297167365791,
"best_model_checkpoint": "./checkpoint-4000",
"epoch": 4.068,
"global_step": 4000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 4.800000000000001e-07,
"loss": 0.689,
"step": 25
},
{
"epoch": 0.01,
"learning_rate": 9.800000000000001e-07,
"loss": 0.4837,
"step": 50
},
{
"epoch": 0.02,
"learning_rate": 1.48e-06,
"loss": 0.3292,
"step": 75
},
{
"epoch": 0.03,
"learning_rate": 1.9600000000000003e-06,
"loss": 0.2195,
"step": 100
},
{
"epoch": 0.03,
"learning_rate": 2.46e-06,
"loss": 0.3355,
"step": 125
},
{
"epoch": 0.04,
"learning_rate": 2.96e-06,
"loss": 0.2592,
"step": 150
},
{
"epoch": 0.04,
"learning_rate": 3.46e-06,
"loss": 0.1808,
"step": 175
},
{
"epoch": 0.05,
"learning_rate": 3.96e-06,
"loss": 0.1772,
"step": 200
},
{
"epoch": 0.06,
"learning_rate": 4.4600000000000005e-06,
"loss": 0.3287,
"step": 225
},
{
"epoch": 0.06,
"learning_rate": 4.960000000000001e-06,
"loss": 0.2778,
"step": 250
},
{
"epoch": 0.07,
"learning_rate": 5.460000000000001e-06,
"loss": 0.1505,
"step": 275
},
{
"epoch": 0.07,
"learning_rate": 5.9600000000000005e-06,
"loss": 0.1308,
"step": 300
},
{
"epoch": 0.08,
"learning_rate": 6.460000000000001e-06,
"loss": 0.1477,
"step": 325
},
{
"epoch": 0.09,
"learning_rate": 6.96e-06,
"loss": 0.1615,
"step": 350
},
{
"epoch": 0.09,
"learning_rate": 7.4600000000000006e-06,
"loss": 0.1905,
"step": 375
},
{
"epoch": 0.1,
"learning_rate": 7.960000000000002e-06,
"loss": 0.2248,
"step": 400
},
{
"epoch": 0.11,
"learning_rate": 8.44e-06,
"loss": 0.2292,
"step": 425
},
{
"epoch": 0.11,
"learning_rate": 8.94e-06,
"loss": 0.1405,
"step": 450
},
{
"epoch": 0.12,
"learning_rate": 9.440000000000001e-06,
"loss": 0.1478,
"step": 475
},
{
"epoch": 0.12,
"learning_rate": 9.940000000000001e-06,
"loss": 0.3437,
"step": 500
},
{
"epoch": 0.13,
"learning_rate": 9.937142857142858e-06,
"loss": 0.204,
"step": 525
},
{
"epoch": 0.14,
"learning_rate": 9.865714285714285e-06,
"loss": 0.2022,
"step": 550
},
{
"epoch": 0.14,
"learning_rate": 9.794285714285714e-06,
"loss": 0.1825,
"step": 575
},
{
"epoch": 0.15,
"learning_rate": 9.722857142857143e-06,
"loss": 0.1732,
"step": 600
},
{
"epoch": 0.16,
"learning_rate": 9.651428571428572e-06,
"loss": 0.1503,
"step": 625
},
{
"epoch": 0.16,
"learning_rate": 9.58e-06,
"loss": 0.1338,
"step": 650
},
{
"epoch": 0.17,
"learning_rate": 9.508571428571429e-06,
"loss": 0.1669,
"step": 675
},
{
"epoch": 0.17,
"learning_rate": 9.437142857142858e-06,
"loss": 0.1137,
"step": 700
},
{
"epoch": 0.18,
"learning_rate": 9.365714285714287e-06,
"loss": 0.1522,
"step": 725
},
{
"epoch": 0.19,
"learning_rate": 9.294285714285714e-06,
"loss": 0.165,
"step": 750
},
{
"epoch": 0.19,
"learning_rate": 9.222857142857143e-06,
"loss": 0.1369,
"step": 775
},
{
"epoch": 0.2,
"learning_rate": 9.151428571428572e-06,
"loss": 0.1367,
"step": 800
},
{
"epoch": 0.21,
"learning_rate": 9.080000000000001e-06,
"loss": 0.116,
"step": 825
},
{
"epoch": 0.21,
"learning_rate": 9.00857142857143e-06,
"loss": 0.1329,
"step": 850
},
{
"epoch": 0.22,
"learning_rate": 8.937142857142857e-06,
"loss": 0.1735,
"step": 875
},
{
"epoch": 0.23,
"learning_rate": 8.865714285714287e-06,
"loss": 0.4067,
"step": 900
},
{
"epoch": 0.23,
"learning_rate": 8.794285714285716e-06,
"loss": 0.1788,
"step": 925
},
{
"epoch": 1.0,
"learning_rate": 8.722857142857145e-06,
"loss": 0.1113,
"step": 950
},
{
"epoch": 1.01,
"learning_rate": 8.651428571428572e-06,
"loss": 0.0608,
"step": 975
},
{
"epoch": 1.02,
"learning_rate": 8.580000000000001e-06,
"loss": 0.1115,
"step": 1000
},
{
"epoch": 1.02,
"eval_loss": 0.2115604728460312,
"eval_runtime": 3420.9371,
"eval_samples_per_second": 0.369,
"eval_steps_per_second": 0.046,
"eval_wer": 9.421690154121695,
"step": 1000
},
{
"epoch": 1.02,
"learning_rate": 8.50857142857143e-06,
"loss": 0.0801,
"step": 1025
},
{
"epoch": 1.03,
"learning_rate": 8.437142857142859e-06,
"loss": 0.0674,
"step": 1050
},
{
"epoch": 1.04,
"learning_rate": 8.365714285714286e-06,
"loss": 0.1268,
"step": 1075
},
{
"epoch": 1.04,
"learning_rate": 8.297142857142859e-06,
"loss": 0.3413,
"step": 1100
},
{
"epoch": 1.05,
"learning_rate": 8.225714285714288e-06,
"loss": 0.1508,
"step": 1125
},
{
"epoch": 1.05,
"learning_rate": 8.154285714285715e-06,
"loss": 0.1074,
"step": 1150
},
{
"epoch": 1.06,
"learning_rate": 8.082857142857144e-06,
"loss": 0.1165,
"step": 1175
},
{
"epoch": 1.07,
"learning_rate": 8.011428571428573e-06,
"loss": 0.1377,
"step": 1200
},
{
"epoch": 1.07,
"learning_rate": 7.94e-06,
"loss": 0.1569,
"step": 1225
},
{
"epoch": 1.08,
"learning_rate": 7.86857142857143e-06,
"loss": 0.0892,
"step": 1250
},
{
"epoch": 1.09,
"learning_rate": 7.797142857142858e-06,
"loss": 0.0844,
"step": 1275
},
{
"epoch": 1.09,
"learning_rate": 7.725714285714286e-06,
"loss": 0.1549,
"step": 1300
},
{
"epoch": 1.1,
"learning_rate": 7.654285714285715e-06,
"loss": 0.1098,
"step": 1325
},
{
"epoch": 1.1,
"learning_rate": 7.5828571428571444e-06,
"loss": 0.0792,
"step": 1350
},
{
"epoch": 1.11,
"learning_rate": 7.511428571428572e-06,
"loss": 0.1332,
"step": 1375
},
{
"epoch": 1.12,
"learning_rate": 7.440000000000001e-06,
"loss": 0.2159,
"step": 1400
},
{
"epoch": 1.12,
"learning_rate": 7.36857142857143e-06,
"loss": 0.1176,
"step": 1425
},
{
"epoch": 1.13,
"learning_rate": 7.297142857142858e-06,
"loss": 0.1513,
"step": 1450
},
{
"epoch": 1.14,
"learning_rate": 7.225714285714286e-06,
"loss": 0.131,
"step": 1475
},
{
"epoch": 1.14,
"learning_rate": 7.154285714285715e-06,
"loss": 0.0901,
"step": 1500
},
{
"epoch": 1.15,
"learning_rate": 7.082857142857143e-06,
"loss": 0.1186,
"step": 1525
},
{
"epoch": 1.15,
"learning_rate": 7.011428571428572e-06,
"loss": 0.1335,
"step": 1550
},
{
"epoch": 1.16,
"learning_rate": 6.9400000000000005e-06,
"loss": 0.0975,
"step": 1575
},
{
"epoch": 1.17,
"learning_rate": 6.868571428571429e-06,
"loss": 0.0927,
"step": 1600
},
{
"epoch": 1.17,
"learning_rate": 6.797142857142858e-06,
"loss": 0.121,
"step": 1625
},
{
"epoch": 1.18,
"learning_rate": 6.728571428571429e-06,
"loss": 0.1175,
"step": 1650
},
{
"epoch": 1.19,
"learning_rate": 6.657142857142857e-06,
"loss": 0.1031,
"step": 1675
},
{
"epoch": 1.19,
"learning_rate": 6.585714285714286e-06,
"loss": 0.1395,
"step": 1700
},
{
"epoch": 1.2,
"learning_rate": 6.514285714285715e-06,
"loss": 0.091,
"step": 1725
},
{
"epoch": 1.2,
"learning_rate": 6.442857142857143e-06,
"loss": 0.0842,
"step": 1750
},
{
"epoch": 1.21,
"learning_rate": 6.371428571428572e-06,
"loss": 0.091,
"step": 1775
},
{
"epoch": 1.22,
"learning_rate": 6.300000000000001e-06,
"loss": 0.1703,
"step": 1800
},
{
"epoch": 1.22,
"learning_rate": 6.22857142857143e-06,
"loss": 0.1302,
"step": 1825
},
{
"epoch": 1.23,
"learning_rate": 6.157142857142858e-06,
"loss": 0.0851,
"step": 1850
},
{
"epoch": 2.0,
"learning_rate": 6.085714285714286e-06,
"loss": 0.0863,
"step": 1875
},
{
"epoch": 2.01,
"learning_rate": 6.014285714285715e-06,
"loss": 0.0645,
"step": 1900
},
{
"epoch": 2.02,
"learning_rate": 5.942857142857143e-06,
"loss": 0.0472,
"step": 1925
},
{
"epoch": 2.02,
"learning_rate": 5.871428571428572e-06,
"loss": 0.062,
"step": 1950
},
{
"epoch": 2.03,
"learning_rate": 5.8e-06,
"loss": 0.1012,
"step": 1975
},
{
"epoch": 2.03,
"learning_rate": 5.7285714285714285e-06,
"loss": 0.0867,
"step": 2000
},
{
"epoch": 2.03,
"eval_loss": 0.19641366600990295,
"eval_runtime": 3448.3792,
"eval_samples_per_second": 0.366,
"eval_steps_per_second": 0.046,
"eval_wer": 9.78225008248103,
"step": 2000
},
{
"epoch": 2.04,
"learning_rate": 5.6571428571428576e-06,
"loss": 0.0951,
"step": 2025
},
{
"epoch": 2.05,
"learning_rate": 5.5857142857142866e-06,
"loss": 0.0759,
"step": 2050
},
{
"epoch": 2.05,
"learning_rate": 5.514285714285714e-06,
"loss": 0.0555,
"step": 2075
},
{
"epoch": 2.06,
"learning_rate": 5.442857142857143e-06,
"loss": 0.0653,
"step": 2100
},
{
"epoch": 2.07,
"learning_rate": 5.371428571428572e-06,
"loss": 0.0822,
"step": 2125
},
{
"epoch": 2.07,
"learning_rate": 5.300000000000001e-06,
"loss": 0.0691,
"step": 2150
},
{
"epoch": 2.08,
"learning_rate": 5.22857142857143e-06,
"loss": 0.0587,
"step": 2175
},
{
"epoch": 2.08,
"learning_rate": 5.157142857142857e-06,
"loss": 0.094,
"step": 2200
},
{
"epoch": 2.09,
"learning_rate": 5.085714285714286e-06,
"loss": 0.0763,
"step": 2225
},
{
"epoch": 2.1,
"learning_rate": 5.014285714285715e-06,
"loss": 0.0516,
"step": 2250
},
{
"epoch": 2.1,
"learning_rate": 4.9428571428571435e-06,
"loss": 0.0393,
"step": 2275
},
{
"epoch": 2.11,
"learning_rate": 4.871428571428572e-06,
"loss": 0.0626,
"step": 2300
},
{
"epoch": 2.12,
"learning_rate": 4.800000000000001e-06,
"loss": 0.1061,
"step": 2325
},
{
"epoch": 2.12,
"learning_rate": 4.728571428571429e-06,
"loss": 0.0622,
"step": 2350
},
{
"epoch": 2.13,
"learning_rate": 4.657142857142857e-06,
"loss": 0.1559,
"step": 2375
},
{
"epoch": 2.13,
"learning_rate": 4.585714285714286e-06,
"loss": 0.1343,
"step": 2400
},
{
"epoch": 2.14,
"learning_rate": 4.514285714285714e-06,
"loss": 0.0636,
"step": 2425
},
{
"epoch": 2.15,
"learning_rate": 4.442857142857143e-06,
"loss": 0.0758,
"step": 2450
},
{
"epoch": 2.15,
"learning_rate": 4.371428571428572e-06,
"loss": 0.0636,
"step": 2475
},
{
"epoch": 2.16,
"learning_rate": 4.3e-06,
"loss": 0.0792,
"step": 2500
},
{
"epoch": 2.17,
"learning_rate": 4.228571428571429e-06,
"loss": 0.094,
"step": 2525
},
{
"epoch": 2.17,
"learning_rate": 4.1571428571428575e-06,
"loss": 0.0967,
"step": 2550
},
{
"epoch": 2.18,
"learning_rate": 4.0857142857142865e-06,
"loss": 0.1417,
"step": 2575
},
{
"epoch": 2.18,
"learning_rate": 4.014285714285715e-06,
"loss": 0.1,
"step": 2600
},
{
"epoch": 2.19,
"learning_rate": 3.942857142857143e-06,
"loss": 0.0721,
"step": 2625
},
{
"epoch": 2.2,
"learning_rate": 3.871428571428572e-06,
"loss": 0.1034,
"step": 2650
},
{
"epoch": 2.2,
"learning_rate": 3.8000000000000005e-06,
"loss": 0.1454,
"step": 2675
},
{
"epoch": 2.21,
"learning_rate": 3.7285714285714286e-06,
"loss": 0.0926,
"step": 2700
},
{
"epoch": 2.22,
"learning_rate": 3.6571428571428576e-06,
"loss": 0.0451,
"step": 2725
},
{
"epoch": 2.22,
"learning_rate": 3.5857142857142862e-06,
"loss": 0.078,
"step": 2750
},
{
"epoch": 2.23,
"learning_rate": 3.5142857142857144e-06,
"loss": 0.0755,
"step": 2775
},
{
"epoch": 3.0,
"learning_rate": 3.4428571428571434e-06,
"loss": 0.0676,
"step": 2800
},
{
"epoch": 3.01,
"learning_rate": 3.3714285714285716e-06,
"loss": 0.0615,
"step": 2825
},
{
"epoch": 3.01,
"learning_rate": 3.3000000000000006e-06,
"loss": 0.0406,
"step": 2850
},
{
"epoch": 3.02,
"learning_rate": 3.2285714285714288e-06,
"loss": 0.0464,
"step": 2875
},
{
"epoch": 3.03,
"learning_rate": 3.1571428571428573e-06,
"loss": 0.0489,
"step": 2900
},
{
"epoch": 3.03,
"learning_rate": 3.085714285714286e-06,
"loss": 0.064,
"step": 2925
},
{
"epoch": 3.04,
"learning_rate": 3.0142857142857145e-06,
"loss": 0.1167,
"step": 2950
},
{
"epoch": 3.04,
"learning_rate": 2.9428571428571427e-06,
"loss": 0.045,
"step": 2975
},
{
"epoch": 3.05,
"learning_rate": 2.8714285714285717e-06,
"loss": 0.0447,
"step": 3000
},
{
"epoch": 3.05,
"eval_loss": 0.20006757974624634,
"eval_runtime": 3456.6956,
"eval_samples_per_second": 0.365,
"eval_steps_per_second": 0.046,
"eval_wer": 9.640854032144036,
"step": 3000
},
{
"epoch": 3.06,
"learning_rate": 2.8000000000000003e-06,
"loss": 0.042,
"step": 3025
},
{
"epoch": 3.06,
"learning_rate": 2.728571428571429e-06,
"loss": 0.0589,
"step": 3050
},
{
"epoch": 3.07,
"learning_rate": 2.6571428571428575e-06,
"loss": 0.065,
"step": 3075
},
{
"epoch": 3.08,
"learning_rate": 2.5857142857142856e-06,
"loss": 0.0409,
"step": 3100
},
{
"epoch": 3.08,
"learning_rate": 2.5142857142857147e-06,
"loss": 0.0513,
"step": 3125
},
{
"epoch": 3.09,
"learning_rate": 2.442857142857143e-06,
"loss": 0.04,
"step": 3150
},
{
"epoch": 3.09,
"learning_rate": 2.371428571428572e-06,
"loss": 0.0409,
"step": 3175
},
{
"epoch": 3.1,
"learning_rate": 2.3000000000000004e-06,
"loss": 0.0461,
"step": 3200
},
{
"epoch": 3.11,
"learning_rate": 2.228571428571429e-06,
"loss": 0.0439,
"step": 3225
},
{
"epoch": 3.11,
"learning_rate": 2.157142857142857e-06,
"loss": 0.0444,
"step": 3250
},
{
"epoch": 3.12,
"learning_rate": 2.0857142857142858e-06,
"loss": 0.057,
"step": 3275
},
{
"epoch": 3.13,
"learning_rate": 2.0142857142857144e-06,
"loss": 0.0353,
"step": 3300
},
{
"epoch": 3.13,
"learning_rate": 1.942857142857143e-06,
"loss": 0.0474,
"step": 3325
},
{
"epoch": 3.14,
"learning_rate": 1.8714285714285715e-06,
"loss": 0.0377,
"step": 3350
},
{
"epoch": 3.14,
"learning_rate": 1.8000000000000001e-06,
"loss": 0.0429,
"step": 3375
},
{
"epoch": 3.15,
"learning_rate": 1.7285714285714287e-06,
"loss": 0.0682,
"step": 3400
},
{
"epoch": 3.16,
"learning_rate": 1.657142857142857e-06,
"loss": 0.0651,
"step": 3425
},
{
"epoch": 3.16,
"learning_rate": 1.5857142857142857e-06,
"loss": 0.1304,
"step": 3450
},
{
"epoch": 3.17,
"learning_rate": 1.5142857142857145e-06,
"loss": 0.0711,
"step": 3475
},
{
"epoch": 3.18,
"learning_rate": 1.442857142857143e-06,
"loss": 0.0658,
"step": 3500
},
{
"epoch": 3.18,
"learning_rate": 1.3714285714285717e-06,
"loss": 0.0478,
"step": 3525
},
{
"epoch": 3.19,
"learning_rate": 1.3e-06,
"loss": 0.0588,
"step": 3550
},
{
"epoch": 3.19,
"learning_rate": 1.2285714285714286e-06,
"loss": 0.0714,
"step": 3575
},
{
"epoch": 3.2,
"learning_rate": 1.1571428571428572e-06,
"loss": 0.0459,
"step": 3600
},
{
"epoch": 3.21,
"learning_rate": 1.0857142857142858e-06,
"loss": 0.0406,
"step": 3625
},
{
"epoch": 3.21,
"learning_rate": 1.0142857142857144e-06,
"loss": 0.0892,
"step": 3650
},
{
"epoch": 3.22,
"learning_rate": 9.42857142857143e-07,
"loss": 0.0855,
"step": 3675
},
{
"epoch": 3.23,
"learning_rate": 8.714285714285716e-07,
"loss": 0.061,
"step": 3700
},
{
"epoch": 3.23,
"learning_rate": 8.000000000000001e-07,
"loss": 0.0458,
"step": 3725
},
{
"epoch": 4.01,
"learning_rate": 7.285714285714287e-07,
"loss": 0.0292,
"step": 3750
},
{
"epoch": 4.01,
"learning_rate": 6.571428571428571e-07,
"loss": 0.04,
"step": 3775
},
{
"epoch": 4.02,
"learning_rate": 5.857142857142857e-07,
"loss": 0.0395,
"step": 3800
},
{
"epoch": 4.02,
"learning_rate": 5.142857142857143e-07,
"loss": 0.0382,
"step": 3825
},
{
"epoch": 4.03,
"learning_rate": 4.4285714285714286e-07,
"loss": 0.1019,
"step": 3850
},
{
"epoch": 4.04,
"learning_rate": 3.7142857142857145e-07,
"loss": 0.0334,
"step": 3875
},
{
"epoch": 4.04,
"learning_rate": 3.0000000000000004e-07,
"loss": 0.04,
"step": 3900
},
{
"epoch": 4.05,
"learning_rate": 2.285714285714286e-07,
"loss": 0.0528,
"step": 3925
},
{
"epoch": 4.06,
"learning_rate": 1.5714285714285717e-07,
"loss": 0.07,
"step": 3950
},
{
"epoch": 4.06,
"learning_rate": 8.571428571428573e-08,
"loss": 0.0406,
"step": 3975
},
{
"epoch": 4.07,
"learning_rate": 1.4285714285714288e-08,
"loss": 0.0426,
"step": 4000
},
{
"epoch": 4.07,
"eval_loss": 0.20508264005184174,
"eval_runtime": 3327.667,
"eval_samples_per_second": 0.379,
"eval_steps_per_second": 0.047,
"eval_wer": 8.335297167365791,
"step": 4000
},
{
"epoch": 4.07,
"step": 4000,
"total_flos": 5.4331946422272e+20,
"train_loss": 0.11459697709977627,
"train_runtime": 90269.9498,
"train_samples_per_second": 2.836,
"train_steps_per_second": 0.044
}
],
"max_steps": 4000,
"num_train_epochs": 9223372036854775807,
"total_flos": 5.4331946422272e+20,
"trial_name": null,
"trial_params": null
}