wav2vec2-bert-CV16-en-libri / trainer_state.json
ylacombe's picture
End of training
92f058b verified
raw
history blame
28.2 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.9962168978562422,
"eval_steps": 250,
"global_step": 1188,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 1.5000000000000002e-08,
"loss": 4.6482,
"step": 5
},
{
"epoch": 0.03,
"learning_rate": 3.0000000000000004e-08,
"loss": 4.7026,
"step": 10
},
{
"epoch": 0.04,
"learning_rate": 4.5e-08,
"loss": 4.6854,
"step": 15
},
{
"epoch": 0.05,
"learning_rate": 6.000000000000001e-08,
"loss": 4.7824,
"step": 20
},
{
"epoch": 0.06,
"learning_rate": 7.500000000000001e-08,
"loss": 4.6708,
"step": 25
},
{
"epoch": 0.08,
"learning_rate": 9e-08,
"loss": 4.6248,
"step": 30
},
{
"epoch": 0.09,
"learning_rate": 1.05e-07,
"loss": 4.7019,
"step": 35
},
{
"epoch": 0.1,
"learning_rate": 1.2000000000000002e-07,
"loss": 4.6747,
"step": 40
},
{
"epoch": 0.11,
"learning_rate": 1.35e-07,
"loss": 4.6209,
"step": 45
},
{
"epoch": 0.13,
"learning_rate": 1.5000000000000002e-07,
"loss": 4.6223,
"step": 50
},
{
"epoch": 0.14,
"learning_rate": 1.6499999999999998e-07,
"loss": 4.5145,
"step": 55
},
{
"epoch": 0.15,
"learning_rate": 1.8e-07,
"loss": 4.5692,
"step": 60
},
{
"epoch": 0.16,
"learning_rate": 1.9499999999999999e-07,
"loss": 4.4948,
"step": 65
},
{
"epoch": 0.18,
"learning_rate": 2.1e-07,
"loss": 4.5019,
"step": 70
},
{
"epoch": 0.19,
"learning_rate": 2.25e-07,
"loss": 4.4425,
"step": 75
},
{
"epoch": 0.2,
"learning_rate": 2.4000000000000003e-07,
"loss": 4.4318,
"step": 80
},
{
"epoch": 0.21,
"learning_rate": 2.5500000000000005e-07,
"loss": 4.428,
"step": 85
},
{
"epoch": 0.23,
"learning_rate": 2.7e-07,
"loss": 4.2983,
"step": 90
},
{
"epoch": 0.24,
"learning_rate": 2.85e-07,
"loss": 4.228,
"step": 95
},
{
"epoch": 0.25,
"learning_rate": 2.97e-07,
"loss": 4.2784,
"step": 100
},
{
"epoch": 0.26,
"learning_rate": 3.12e-07,
"loss": 4.1254,
"step": 105
},
{
"epoch": 0.28,
"learning_rate": 3.27e-07,
"loss": 4.0566,
"step": 110
},
{
"epoch": 0.29,
"learning_rate": 3.42e-07,
"loss": 4.0241,
"step": 115
},
{
"epoch": 0.3,
"learning_rate": 3.5700000000000003e-07,
"loss": 4.0261,
"step": 120
},
{
"epoch": 0.32,
"learning_rate": 3.72e-07,
"loss": 3.864,
"step": 125
},
{
"epoch": 0.33,
"learning_rate": 3.87e-07,
"loss": 3.8068,
"step": 130
},
{
"epoch": 0.34,
"learning_rate": 4.02e-07,
"loss": 3.7553,
"step": 135
},
{
"epoch": 0.35,
"learning_rate": 4.17e-07,
"loss": 3.6628,
"step": 140
},
{
"epoch": 0.37,
"learning_rate": 4.32e-07,
"loss": 3.5843,
"step": 145
},
{
"epoch": 0.38,
"learning_rate": 4.47e-07,
"loss": 3.5552,
"step": 150
},
{
"epoch": 0.39,
"learning_rate": 4.6200000000000003e-07,
"loss": 3.4307,
"step": 155
},
{
"epoch": 0.4,
"learning_rate": 4.77e-07,
"loss": 3.3634,
"step": 160
},
{
"epoch": 0.42,
"learning_rate": 4.92e-07,
"loss": 3.3041,
"step": 165
},
{
"epoch": 0.43,
"learning_rate": 5.07e-07,
"loss": 3.2511,
"step": 170
},
{
"epoch": 0.44,
"learning_rate": 5.219999999999999e-07,
"loss": 3.1583,
"step": 175
},
{
"epoch": 0.45,
"learning_rate": 5.37e-07,
"loss": 3.1013,
"step": 180
},
{
"epoch": 0.47,
"learning_rate": 5.52e-07,
"loss": 3.0769,
"step": 185
},
{
"epoch": 0.48,
"learning_rate": 5.67e-07,
"loss": 3.0122,
"step": 190
},
{
"epoch": 0.49,
"learning_rate": 5.82e-07,
"loss": 2.9856,
"step": 195
},
{
"epoch": 0.5,
"learning_rate": 5.970000000000001e-07,
"loss": 2.9903,
"step": 200
},
{
"epoch": 0.52,
"learning_rate": 6.12e-07,
"loss": 2.9308,
"step": 205
},
{
"epoch": 0.53,
"learning_rate": 6.27e-07,
"loss": 2.9159,
"step": 210
},
{
"epoch": 0.54,
"learning_rate": 6.42e-07,
"loss": 2.9075,
"step": 215
},
{
"epoch": 0.55,
"learning_rate": 6.57e-07,
"loss": 2.8958,
"step": 220
},
{
"epoch": 0.57,
"learning_rate": 6.72e-07,
"loss": 2.8698,
"step": 225
},
{
"epoch": 0.58,
"learning_rate": 6.87e-07,
"loss": 2.8697,
"step": 230
},
{
"epoch": 0.59,
"learning_rate": 7.02e-07,
"loss": 2.8911,
"step": 235
},
{
"epoch": 0.61,
"learning_rate": 7.170000000000001e-07,
"loss": 2.8506,
"step": 240
},
{
"epoch": 0.62,
"learning_rate": 7.32e-07,
"loss": 2.8522,
"step": 245
},
{
"epoch": 0.63,
"learning_rate": 7.47e-07,
"loss": 2.8812,
"step": 250
},
{
"epoch": 0.63,
"eval_cer": 0.9999960063100302,
"eval_loss": 2.892289161682129,
"eval_runtime": 59.7132,
"eval_samples_per_second": 42.336,
"eval_steps_per_second": 1.189,
"eval_wer": 1.0,
"step": 250
},
{
"epoch": 0.64,
"learning_rate": 7.62e-07,
"loss": 2.8429,
"step": 255
},
{
"epoch": 0.66,
"learning_rate": 7.77e-07,
"loss": 2.8424,
"step": 260
},
{
"epoch": 0.67,
"learning_rate": 7.92e-07,
"loss": 2.8437,
"step": 265
},
{
"epoch": 0.68,
"learning_rate": 8.070000000000001e-07,
"loss": 2.8541,
"step": 270
},
{
"epoch": 0.69,
"learning_rate": 8.22e-07,
"loss": 2.8411,
"step": 275
},
{
"epoch": 0.71,
"learning_rate": 8.370000000000001e-07,
"loss": 2.8387,
"step": 280
},
{
"epoch": 0.72,
"learning_rate": 8.520000000000001e-07,
"loss": 2.8561,
"step": 285
},
{
"epoch": 0.73,
"learning_rate": 8.67e-07,
"loss": 2.8254,
"step": 290
},
{
"epoch": 0.74,
"learning_rate": 8.82e-07,
"loss": 2.8244,
"step": 295
},
{
"epoch": 0.76,
"learning_rate": 8.97e-07,
"loss": 2.8409,
"step": 300
},
{
"epoch": 0.77,
"learning_rate": 9.12e-07,
"loss": 2.816,
"step": 305
},
{
"epoch": 0.78,
"learning_rate": 9.270000000000001e-07,
"loss": 2.8173,
"step": 310
},
{
"epoch": 0.79,
"learning_rate": 9.419999999999999e-07,
"loss": 2.8253,
"step": 315
},
{
"epoch": 0.81,
"learning_rate": 9.57e-07,
"loss": 2.8185,
"step": 320
},
{
"epoch": 0.82,
"learning_rate": 9.72e-07,
"loss": 2.8046,
"step": 325
},
{
"epoch": 0.83,
"learning_rate": 9.87e-07,
"loss": 2.8048,
"step": 330
},
{
"epoch": 0.84,
"learning_rate": 1.002e-06,
"loss": 2.8177,
"step": 335
},
{
"epoch": 0.86,
"learning_rate": 1.017e-06,
"loss": 2.775,
"step": 340
},
{
"epoch": 0.87,
"learning_rate": 1.032e-06,
"loss": 2.7762,
"step": 345
},
{
"epoch": 0.88,
"learning_rate": 1.047e-06,
"loss": 2.7735,
"step": 350
},
{
"epoch": 0.9,
"learning_rate": 1.062e-06,
"loss": 2.7319,
"step": 355
},
{
"epoch": 0.91,
"learning_rate": 1.077e-06,
"loss": 2.71,
"step": 360
},
{
"epoch": 0.92,
"learning_rate": 1.092e-06,
"loss": 2.6868,
"step": 365
},
{
"epoch": 0.93,
"learning_rate": 1.1070000000000002e-06,
"loss": 2.6709,
"step": 370
},
{
"epoch": 0.95,
"learning_rate": 1.122e-06,
"loss": 2.6141,
"step": 375
},
{
"epoch": 0.96,
"learning_rate": 1.137e-06,
"loss": 2.5824,
"step": 380
},
{
"epoch": 0.97,
"learning_rate": 1.152e-06,
"loss": 2.5791,
"step": 385
},
{
"epoch": 0.98,
"learning_rate": 1.167e-06,
"loss": 2.5124,
"step": 390
},
{
"epoch": 1.0,
"learning_rate": 1.179e-06,
"loss": 2.4623,
"step": 395
},
{
"epoch": 1.01,
"learning_rate": 1.1940000000000001e-06,
"loss": 2.4341,
"step": 400
},
{
"epoch": 1.02,
"learning_rate": 1.2090000000000002e-06,
"loss": 2.3441,
"step": 405
},
{
"epoch": 1.03,
"learning_rate": 1.224e-06,
"loss": 2.2861,
"step": 410
},
{
"epoch": 1.05,
"learning_rate": 1.2390000000000001e-06,
"loss": 2.2725,
"step": 415
},
{
"epoch": 1.06,
"learning_rate": 1.254e-06,
"loss": 2.1615,
"step": 420
},
{
"epoch": 1.07,
"learning_rate": 1.269e-06,
"loss": 2.0869,
"step": 425
},
{
"epoch": 1.08,
"learning_rate": 1.284e-06,
"loss": 2.0713,
"step": 430
},
{
"epoch": 1.1,
"learning_rate": 1.299e-06,
"loss": 1.9363,
"step": 435
},
{
"epoch": 1.11,
"learning_rate": 1.314e-06,
"loss": 1.8633,
"step": 440
},
{
"epoch": 1.12,
"learning_rate": 1.3290000000000001e-06,
"loss": 1.8307,
"step": 445
},
{
"epoch": 1.13,
"learning_rate": 1.344e-06,
"loss": 1.7683,
"step": 450
},
{
"epoch": 1.15,
"learning_rate": 1.359e-06,
"loss": 1.6742,
"step": 455
},
{
"epoch": 1.16,
"learning_rate": 1.374e-06,
"loss": 1.6215,
"step": 460
},
{
"epoch": 1.17,
"learning_rate": 1.389e-06,
"loss": 1.6537,
"step": 465
},
{
"epoch": 1.19,
"learning_rate": 1.404e-06,
"loss": 1.5039,
"step": 470
},
{
"epoch": 1.2,
"learning_rate": 1.419e-06,
"loss": 1.4679,
"step": 475
},
{
"epoch": 1.21,
"learning_rate": 1.4340000000000002e-06,
"loss": 1.4565,
"step": 480
},
{
"epoch": 1.22,
"learning_rate": 1.449e-06,
"loss": 1.3904,
"step": 485
},
{
"epoch": 1.24,
"learning_rate": 1.464e-06,
"loss": 1.3057,
"step": 490
},
{
"epoch": 1.25,
"learning_rate": 1.479e-06,
"loss": 1.2998,
"step": 495
},
{
"epoch": 1.26,
"learning_rate": 1.494e-06,
"loss": 1.2899,
"step": 500
},
{
"epoch": 1.26,
"eval_cer": 0.2562710916751533,
"eval_loss": 1.147094488143921,
"eval_runtime": 60.746,
"eval_samples_per_second": 41.616,
"eval_steps_per_second": 1.169,
"eval_wer": 0.7029631516386483,
"step": 500
},
{
"epoch": 1.27,
"learning_rate": 1.5089999999999999e-06,
"loss": 1.2025,
"step": 505
},
{
"epoch": 1.29,
"learning_rate": 1.524e-06,
"loss": 1.1523,
"step": 510
},
{
"epoch": 1.3,
"learning_rate": 1.539e-06,
"loss": 1.1846,
"step": 515
},
{
"epoch": 1.31,
"learning_rate": 1.554e-06,
"loss": 1.0974,
"step": 520
},
{
"epoch": 1.32,
"learning_rate": 1.569e-06,
"loss": 1.0635,
"step": 525
},
{
"epoch": 1.34,
"learning_rate": 1.584e-06,
"loss": 1.0922,
"step": 530
},
{
"epoch": 1.35,
"learning_rate": 1.599e-06,
"loss": 1.0167,
"step": 535
},
{
"epoch": 1.36,
"learning_rate": 1.6140000000000001e-06,
"loss": 0.9946,
"step": 540
},
{
"epoch": 1.37,
"learning_rate": 1.629e-06,
"loss": 1.0021,
"step": 545
},
{
"epoch": 1.39,
"learning_rate": 1.644e-06,
"loss": 1.006,
"step": 550
},
{
"epoch": 1.4,
"learning_rate": 1.6590000000000001e-06,
"loss": 0.9312,
"step": 555
},
{
"epoch": 1.41,
"learning_rate": 1.6740000000000002e-06,
"loss": 0.9275,
"step": 560
},
{
"epoch": 1.42,
"learning_rate": 1.689e-06,
"loss": 0.9743,
"step": 565
},
{
"epoch": 1.44,
"learning_rate": 1.7040000000000001e-06,
"loss": 0.8707,
"step": 570
},
{
"epoch": 1.45,
"learning_rate": 1.719e-06,
"loss": 0.8765,
"step": 575
},
{
"epoch": 1.46,
"learning_rate": 1.734e-06,
"loss": 0.9403,
"step": 580
},
{
"epoch": 1.48,
"learning_rate": 1.749e-06,
"loss": 0.8242,
"step": 585
},
{
"epoch": 1.49,
"learning_rate": 1.764e-06,
"loss": 0.7783,
"step": 590
},
{
"epoch": 1.5,
"learning_rate": 1.779e-06,
"loss": 0.8126,
"step": 595
},
{
"epoch": 1.51,
"learning_rate": 1.794e-06,
"loss": 0.8544,
"step": 600
},
{
"epoch": 1.53,
"learning_rate": 1.809e-06,
"loss": 0.7637,
"step": 605
},
{
"epoch": 1.54,
"learning_rate": 1.824e-06,
"loss": 0.7481,
"step": 610
},
{
"epoch": 1.55,
"learning_rate": 1.839e-06,
"loss": 0.8198,
"step": 615
},
{
"epoch": 1.56,
"learning_rate": 1.8540000000000002e-06,
"loss": 0.7086,
"step": 620
},
{
"epoch": 1.58,
"learning_rate": 1.869e-06,
"loss": 0.7167,
"step": 625
},
{
"epoch": 1.59,
"learning_rate": 1.8810000000000003e-06,
"loss": 0.7675,
"step": 630
},
{
"epoch": 1.6,
"learning_rate": 1.8960000000000001e-06,
"loss": 0.6906,
"step": 635
},
{
"epoch": 1.61,
"learning_rate": 1.9110000000000004e-06,
"loss": 0.6739,
"step": 640
},
{
"epoch": 1.63,
"learning_rate": 1.926e-06,
"loss": 0.7041,
"step": 645
},
{
"epoch": 1.64,
"learning_rate": 1.9409999999999997e-06,
"loss": 0.6874,
"step": 650
},
{
"epoch": 1.65,
"learning_rate": 1.956e-06,
"loss": 0.6738,
"step": 655
},
{
"epoch": 1.66,
"learning_rate": 1.971e-06,
"loss": 0.6507,
"step": 660
},
{
"epoch": 1.68,
"learning_rate": 1.9859999999999997e-06,
"loss": 0.6857,
"step": 665
},
{
"epoch": 1.69,
"learning_rate": 2.001e-06,
"loss": 0.6291,
"step": 670
},
{
"epoch": 1.7,
"learning_rate": 2.016e-06,
"loss": 0.5821,
"step": 675
},
{
"epoch": 1.72,
"learning_rate": 2.031e-06,
"loss": 0.6688,
"step": 680
},
{
"epoch": 1.73,
"learning_rate": 2.046e-06,
"loss": 0.5958,
"step": 685
},
{
"epoch": 1.74,
"learning_rate": 2.061e-06,
"loss": 0.5807,
"step": 690
},
{
"epoch": 1.75,
"learning_rate": 2.076e-06,
"loss": 0.5991,
"step": 695
},
{
"epoch": 1.77,
"learning_rate": 2.091e-06,
"loss": 0.5904,
"step": 700
},
{
"epoch": 1.78,
"learning_rate": 2.106e-06,
"loss": 0.5715,
"step": 705
},
{
"epoch": 1.79,
"learning_rate": 2.121e-06,
"loss": 0.563,
"step": 710
},
{
"epoch": 1.8,
"learning_rate": 2.136e-06,
"loss": 0.6291,
"step": 715
},
{
"epoch": 1.82,
"learning_rate": 2.1510000000000002e-06,
"loss": 0.5513,
"step": 720
},
{
"epoch": 1.83,
"learning_rate": 2.166e-06,
"loss": 0.539,
"step": 725
},
{
"epoch": 1.84,
"learning_rate": 2.181e-06,
"loss": 0.6234,
"step": 730
},
{
"epoch": 1.85,
"learning_rate": 2.1960000000000002e-06,
"loss": 0.5196,
"step": 735
},
{
"epoch": 1.87,
"learning_rate": 2.211e-06,
"loss": 0.5122,
"step": 740
},
{
"epoch": 1.88,
"learning_rate": 2.226e-06,
"loss": 0.5331,
"step": 745
},
{
"epoch": 1.89,
"learning_rate": 2.2410000000000002e-06,
"loss": 0.5276,
"step": 750
},
{
"epoch": 1.89,
"eval_cer": 0.11271790570897981,
"eval_loss": 0.46866554021835327,
"eval_runtime": 59.8626,
"eval_samples_per_second": 42.23,
"eval_steps_per_second": 1.186,
"eval_wer": 0.41144506707420614,
"step": 750
},
{
"epoch": 1.9,
"learning_rate": 2.256e-06,
"loss": 0.49,
"step": 755
},
{
"epoch": 1.92,
"learning_rate": 2.2710000000000004e-06,
"loss": 0.4961,
"step": 760
},
{
"epoch": 1.93,
"learning_rate": 2.2860000000000002e-06,
"loss": 0.5207,
"step": 765
},
{
"epoch": 1.94,
"learning_rate": 2.301e-06,
"loss": 0.4767,
"step": 770
},
{
"epoch": 1.95,
"learning_rate": 2.3160000000000004e-06,
"loss": 0.4761,
"step": 775
},
{
"epoch": 1.97,
"learning_rate": 2.3310000000000002e-06,
"loss": 0.5454,
"step": 780
},
{
"epoch": 1.98,
"learning_rate": 2.346e-06,
"loss": 0.4688,
"step": 785
},
{
"epoch": 1.99,
"learning_rate": 2.3610000000000003e-06,
"loss": 0.4609,
"step": 790
},
{
"epoch": 2.01,
"learning_rate": 2.376e-06,
"loss": 0.4967,
"step": 795
},
{
"epoch": 2.02,
"learning_rate": 2.391e-06,
"loss": 0.4613,
"step": 800
},
{
"epoch": 2.03,
"learning_rate": 2.406e-06,
"loss": 0.448,
"step": 805
},
{
"epoch": 2.04,
"learning_rate": 2.4209999999999998e-06,
"loss": 0.5151,
"step": 810
},
{
"epoch": 2.06,
"learning_rate": 2.436e-06,
"loss": 0.4411,
"step": 815
},
{
"epoch": 2.07,
"learning_rate": 2.451e-06,
"loss": 0.43,
"step": 820
},
{
"epoch": 2.08,
"learning_rate": 2.4659999999999998e-06,
"loss": 0.4318,
"step": 825
},
{
"epoch": 2.09,
"learning_rate": 2.481e-06,
"loss": 0.4442,
"step": 830
},
{
"epoch": 2.11,
"learning_rate": 2.496e-06,
"loss": 0.4014,
"step": 835
},
{
"epoch": 2.12,
"learning_rate": 2.5109999999999998e-06,
"loss": 0.4062,
"step": 840
},
{
"epoch": 2.13,
"learning_rate": 2.526e-06,
"loss": 0.4431,
"step": 845
},
{
"epoch": 2.14,
"learning_rate": 2.541e-06,
"loss": 0.4057,
"step": 850
},
{
"epoch": 2.16,
"learning_rate": 2.556e-06,
"loss": 0.3873,
"step": 855
},
{
"epoch": 2.17,
"learning_rate": 2.571e-06,
"loss": 0.4497,
"step": 860
},
{
"epoch": 2.18,
"learning_rate": 2.586e-06,
"loss": 0.3917,
"step": 865
},
{
"epoch": 2.19,
"learning_rate": 2.601e-06,
"loss": 0.363,
"step": 870
},
{
"epoch": 2.21,
"learning_rate": 2.616e-06,
"loss": 0.4173,
"step": 875
},
{
"epoch": 2.22,
"learning_rate": 2.631e-06,
"loss": 0.4128,
"step": 880
},
{
"epoch": 2.23,
"learning_rate": 2.646e-06,
"loss": 0.3761,
"step": 885
},
{
"epoch": 2.24,
"learning_rate": 2.661e-06,
"loss": 0.3817,
"step": 890
},
{
"epoch": 2.26,
"learning_rate": 2.6760000000000003e-06,
"loss": 0.4256,
"step": 895
},
{
"epoch": 2.27,
"learning_rate": 2.691e-06,
"loss": 0.3778,
"step": 900
},
{
"epoch": 2.28,
"learning_rate": 2.706e-06,
"loss": 0.3571,
"step": 905
},
{
"epoch": 2.3,
"learning_rate": 2.7210000000000003e-06,
"loss": 0.4345,
"step": 910
},
{
"epoch": 2.31,
"learning_rate": 2.736e-06,
"loss": 0.3557,
"step": 915
},
{
"epoch": 2.32,
"learning_rate": 2.751e-06,
"loss": 0.3616,
"step": 920
},
{
"epoch": 2.33,
"learning_rate": 2.7660000000000003e-06,
"loss": 0.3813,
"step": 925
},
{
"epoch": 2.35,
"learning_rate": 2.781e-06,
"loss": 0.3828,
"step": 930
},
{
"epoch": 2.36,
"learning_rate": 2.7960000000000004e-06,
"loss": 0.3234,
"step": 935
},
{
"epoch": 2.37,
"learning_rate": 2.8110000000000003e-06,
"loss": 0.3353,
"step": 940
},
{
"epoch": 2.38,
"learning_rate": 2.826e-06,
"loss": 0.3856,
"step": 945
},
{
"epoch": 2.4,
"learning_rate": 2.8410000000000004e-06,
"loss": 0.3371,
"step": 950
},
{
"epoch": 2.41,
"learning_rate": 2.8560000000000003e-06,
"loss": 0.316,
"step": 955
},
{
"epoch": 2.42,
"learning_rate": 2.8709999999999997e-06,
"loss": 0.3904,
"step": 960
},
{
"epoch": 2.43,
"learning_rate": 2.886e-06,
"loss": 0.3297,
"step": 965
},
{
"epoch": 2.45,
"learning_rate": 2.901e-06,
"loss": 0.3163,
"step": 970
},
{
"epoch": 2.46,
"learning_rate": 2.916e-06,
"loss": 0.3396,
"step": 975
},
{
"epoch": 2.47,
"learning_rate": 2.931e-06,
"loss": 0.3422,
"step": 980
},
{
"epoch": 2.48,
"learning_rate": 2.946e-06,
"loss": 0.3075,
"step": 985
},
{
"epoch": 2.5,
"learning_rate": 2.961e-06,
"loss": 0.3132,
"step": 990
},
{
"epoch": 2.51,
"learning_rate": 2.976e-06,
"loss": 0.3504,
"step": 995
},
{
"epoch": 2.52,
"learning_rate": 2.991e-06,
"loss": 0.3313,
"step": 1000
},
{
"epoch": 2.52,
"eval_cer": 0.065871922362667,
"eval_loss": 0.287024587392807,
"eval_runtime": 60.4937,
"eval_samples_per_second": 41.789,
"eval_steps_per_second": 1.174,
"eval_wer": 0.2576838172864663,
"step": 1000
},
{
"epoch": 2.53,
"learning_rate": 3.006e-06,
"loss": 0.306,
"step": 1005
},
{
"epoch": 2.55,
"learning_rate": 3.021e-06,
"loss": 0.3832,
"step": 1010
},
{
"epoch": 2.56,
"learning_rate": 3.036e-06,
"loss": 0.2945,
"step": 1015
},
{
"epoch": 2.57,
"learning_rate": 3.051e-06,
"loss": 0.2986,
"step": 1020
},
{
"epoch": 2.59,
"learning_rate": 3.066e-06,
"loss": 0.328,
"step": 1025
},
{
"epoch": 2.6,
"learning_rate": 3.0810000000000002e-06,
"loss": 0.341,
"step": 1030
},
{
"epoch": 2.61,
"learning_rate": 3.096e-06,
"loss": 0.2928,
"step": 1035
},
{
"epoch": 2.62,
"learning_rate": 3.111e-06,
"loss": 0.3032,
"step": 1040
},
{
"epoch": 2.64,
"learning_rate": 3.1260000000000002e-06,
"loss": 0.3132,
"step": 1045
},
{
"epoch": 2.65,
"learning_rate": 3.141e-06,
"loss": 0.2795,
"step": 1050
},
{
"epoch": 2.66,
"learning_rate": 3.156e-06,
"loss": 0.3074,
"step": 1055
},
{
"epoch": 2.67,
"learning_rate": 3.1710000000000002e-06,
"loss": 0.3546,
"step": 1060
},
{
"epoch": 2.69,
"learning_rate": 3.186e-06,
"loss": 0.2661,
"step": 1065
},
{
"epoch": 2.7,
"learning_rate": 3.2010000000000004e-06,
"loss": 0.2756,
"step": 1070
},
{
"epoch": 2.71,
"learning_rate": 3.216e-06,
"loss": 0.3041,
"step": 1075
},
{
"epoch": 2.72,
"learning_rate": 3.231e-06,
"loss": 0.3163,
"step": 1080
},
{
"epoch": 2.74,
"learning_rate": 3.2460000000000003e-06,
"loss": 0.2646,
"step": 1085
},
{
"epoch": 2.75,
"learning_rate": 3.261e-06,
"loss": 0.2772,
"step": 1090
},
{
"epoch": 2.76,
"learning_rate": 3.276e-06,
"loss": 0.3198,
"step": 1095
},
{
"epoch": 2.77,
"learning_rate": 3.2910000000000003e-06,
"loss": 0.2586,
"step": 1100
},
{
"epoch": 2.79,
"learning_rate": 3.306e-06,
"loss": 0.2671,
"step": 1105
},
{
"epoch": 2.8,
"learning_rate": 3.3210000000000005e-06,
"loss": 0.3135,
"step": 1110
},
{
"epoch": 2.81,
"learning_rate": 3.336e-06,
"loss": 0.2482,
"step": 1115
},
{
"epoch": 2.82,
"learning_rate": 3.3509999999999998e-06,
"loss": 0.2427,
"step": 1120
},
{
"epoch": 2.84,
"learning_rate": 3.366e-06,
"loss": 0.2821,
"step": 1125
},
{
"epoch": 2.85,
"learning_rate": 3.381e-06,
"loss": 0.2651,
"step": 1130
},
{
"epoch": 2.86,
"learning_rate": 3.3959999999999998e-06,
"loss": 0.2517,
"step": 1135
},
{
"epoch": 2.88,
"learning_rate": 3.411e-06,
"loss": 0.2528,
"step": 1140
},
{
"epoch": 2.89,
"learning_rate": 3.426e-06,
"loss": 0.3223,
"step": 1145
},
{
"epoch": 2.9,
"learning_rate": 3.441e-06,
"loss": 0.2383,
"step": 1150
},
{
"epoch": 2.91,
"learning_rate": 3.456e-06,
"loss": 0.2402,
"step": 1155
},
{
"epoch": 2.93,
"learning_rate": 3.471e-06,
"loss": 0.2866,
"step": 1160
},
{
"epoch": 2.94,
"learning_rate": 3.486e-06,
"loss": 0.2316,
"step": 1165
},
{
"epoch": 2.95,
"learning_rate": 3.501e-06,
"loss": 0.2293,
"step": 1170
},
{
"epoch": 2.96,
"learning_rate": 3.516e-06,
"loss": 0.2607,
"step": 1175
},
{
"epoch": 2.98,
"learning_rate": 3.531e-06,
"loss": 0.2492,
"step": 1180
},
{
"epoch": 2.99,
"learning_rate": 3.546e-06,
"loss": 0.2366,
"step": 1185
},
{
"epoch": 3.0,
"step": 1188,
"total_flos": 3.172956745754896e+19,
"train_loss": 1.5876106701113961,
"train_runtime": 5725.9395,
"train_samples_per_second": 14.952,
"train_steps_per_second": 0.207
}
],
"logging_steps": 5,
"max_steps": 1188,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"total_flos": 3.172956745754896e+19,
"train_batch_size": 12,
"trial_name": null,
"trial_params": null
}