sew-small-100k-timit / trainer_state.json
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 20.0,
"global_step": 2900,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.07,
"learning_rate": 8.000000000000001e-07,
"loss": 9.5879,
"step": 10
},
{
"epoch": 0.14,
"learning_rate": 1.8e-06,
"loss": 8.4067,
"step": 20
},
{
"epoch": 0.21,
"learning_rate": 2.8000000000000003e-06,
"loss": 7.8507,
"step": 30
},
{
"epoch": 0.28,
"learning_rate": 3.7e-06,
"loss": 8.4293,
"step": 40
},
{
"epoch": 0.34,
"learning_rate": 4.7e-06,
"loss": 7.6196,
"step": 50
},
{
"epoch": 0.41,
"learning_rate": 5.7000000000000005e-06,
"loss": 6.565,
"step": 60
},
{
"epoch": 0.48,
"learning_rate": 6.700000000000001e-06,
"loss": 5.0036,
"step": 70
},
{
"epoch": 0.55,
"learning_rate": 7.7e-06,
"loss": 3.9251,
"step": 80
},
{
"epoch": 0.62,
"learning_rate": 8.7e-06,
"loss": 3.2412,
"step": 90
},
{
"epoch": 0.69,
"learning_rate": 9.7e-06,
"loss": 3.071,
"step": 100
},
{
"epoch": 0.69,
"eval_loss": 3.0262322425842285,
"eval_runtime": 53.4663,
"eval_samples_per_second": 31.422,
"eval_steps_per_second": 31.422,
"eval_wer": 1.0,
"step": 100
},
{
"epoch": 0.76,
"learning_rate": 1.0700000000000001e-05,
"loss": 3.0293,
"step": 110
},
{
"epoch": 0.83,
"learning_rate": 1.1700000000000001e-05,
"loss": 2.9589,
"step": 120
},
{
"epoch": 0.9,
"learning_rate": 1.27e-05,
"loss": 2.9437,
"step": 130
},
{
"epoch": 0.97,
"learning_rate": 1.3700000000000001e-05,
"loss": 2.9449,
"step": 140
},
{
"epoch": 1.03,
"learning_rate": 1.47e-05,
"loss": 2.9699,
"step": 150
},
{
"epoch": 1.1,
"learning_rate": 1.5700000000000002e-05,
"loss": 2.9409,
"step": 160
},
{
"epoch": 1.17,
"learning_rate": 1.6700000000000003e-05,
"loss": 2.9301,
"step": 170
},
{
"epoch": 1.24,
"learning_rate": 1.77e-05,
"loss": 2.9621,
"step": 180
},
{
"epoch": 1.31,
"learning_rate": 1.87e-05,
"loss": 2.95,
"step": 190
},
{
"epoch": 1.38,
"learning_rate": 1.97e-05,
"loss": 2.9304,
"step": 200
},
{
"epoch": 1.38,
"eval_loss": 2.92971134185791,
"eval_runtime": 53.2709,
"eval_samples_per_second": 31.537,
"eval_steps_per_second": 31.537,
"eval_wer": 1.0,
"step": 200
},
{
"epoch": 1.45,
"learning_rate": 2.07e-05,
"loss": 2.934,
"step": 210
},
{
"epoch": 1.52,
"learning_rate": 2.1700000000000002e-05,
"loss": 2.934,
"step": 220
},
{
"epoch": 1.59,
"learning_rate": 2.2700000000000003e-05,
"loss": 2.9189,
"step": 230
},
{
"epoch": 1.66,
"learning_rate": 2.37e-05,
"loss": 2.9029,
"step": 240
},
{
"epoch": 1.72,
"learning_rate": 2.47e-05,
"loss": 2.9062,
"step": 250
},
{
"epoch": 1.79,
"learning_rate": 2.57e-05,
"loss": 2.9382,
"step": 260
},
{
"epoch": 1.86,
"learning_rate": 2.6700000000000002e-05,
"loss": 2.8572,
"step": 270
},
{
"epoch": 1.93,
"learning_rate": 2.7700000000000002e-05,
"loss": 2.8821,
"step": 280
},
{
"epoch": 2.0,
"learning_rate": 2.87e-05,
"loss": 2.9294,
"step": 290
},
{
"epoch": 2.07,
"learning_rate": 2.97e-05,
"loss": 2.8823,
"step": 300
},
{
"epoch": 2.07,
"eval_loss": 2.8366596698760986,
"eval_runtime": 53.4753,
"eval_samples_per_second": 31.416,
"eval_steps_per_second": 31.416,
"eval_wer": 1.0,
"step": 300
},
{
"epoch": 2.14,
"learning_rate": 3.07e-05,
"loss": 2.7743,
"step": 310
},
{
"epoch": 2.21,
"learning_rate": 3.1700000000000005e-05,
"loss": 2.7128,
"step": 320
},
{
"epoch": 2.28,
"learning_rate": 3.27e-05,
"loss": 2.6753,
"step": 330
},
{
"epoch": 2.34,
"learning_rate": 3.3700000000000006e-05,
"loss": 2.4534,
"step": 340
},
{
"epoch": 2.41,
"learning_rate": 3.4699999999999996e-05,
"loss": 2.3182,
"step": 350
},
{
"epoch": 2.48,
"learning_rate": 3.57e-05,
"loss": 2.1757,
"step": 360
},
{
"epoch": 2.55,
"learning_rate": 3.6700000000000004e-05,
"loss": 2.0205,
"step": 370
},
{
"epoch": 2.62,
"learning_rate": 3.77e-05,
"loss": 1.7001,
"step": 380
},
{
"epoch": 2.69,
"learning_rate": 3.8700000000000006e-05,
"loss": 1.5838,
"step": 390
},
{
"epoch": 2.76,
"learning_rate": 3.97e-05,
"loss": 1.5668,
"step": 400
},
{
"epoch": 2.76,
"eval_loss": 1.230989933013916,
"eval_runtime": 53.6866,
"eval_samples_per_second": 31.293,
"eval_steps_per_second": 31.293,
"eval_wer": 0.8807111846185652,
"step": 400
},
{
"epoch": 2.83,
"learning_rate": 4.07e-05,
"loss": 1.2143,
"step": 410
},
{
"epoch": 2.9,
"learning_rate": 4.17e-05,
"loss": 1.06,
"step": 420
},
{
"epoch": 2.97,
"learning_rate": 4.27e-05,
"loss": 1.0786,
"step": 430
},
{
"epoch": 3.03,
"learning_rate": 4.3700000000000005e-05,
"loss": 1.1552,
"step": 440
},
{
"epoch": 3.1,
"learning_rate": 4.47e-05,
"loss": 0.8438,
"step": 450
},
{
"epoch": 3.17,
"learning_rate": 4.5700000000000006e-05,
"loss": 0.8164,
"step": 460
},
{
"epoch": 3.24,
"learning_rate": 4.6700000000000003e-05,
"loss": 0.927,
"step": 470
},
{
"epoch": 3.31,
"learning_rate": 4.77e-05,
"loss": 0.8643,
"step": 480
},
{
"epoch": 3.38,
"learning_rate": 4.87e-05,
"loss": 0.7093,
"step": 490
},
{
"epoch": 3.45,
"learning_rate": 4.97e-05,
"loss": 0.7422,
"step": 500
},
{
"epoch": 3.45,
"eval_loss": 0.7080259919166565,
"eval_runtime": 54.074,
"eval_samples_per_second": 31.069,
"eval_steps_per_second": 31.069,
"eval_wer": 0.5956860312866101,
"step": 500
},
{
"epoch": 3.52,
"learning_rate": 5.0700000000000006e-05,
"loss": 0.8921,
"step": 510
},
{
"epoch": 3.59,
"learning_rate": 5.17e-05,
"loss": 0.6304,
"step": 520
},
{
"epoch": 3.66,
"learning_rate": 5.270000000000001e-05,
"loss": 0.6065,
"step": 530
},
{
"epoch": 3.72,
"learning_rate": 5.3700000000000004e-05,
"loss": 0.754,
"step": 540
},
{
"epoch": 3.79,
"learning_rate": 5.470000000000001e-05,
"loss": 0.7717,
"step": 550
},
{
"epoch": 3.86,
"learning_rate": 5.5700000000000005e-05,
"loss": 0.5817,
"step": 560
},
{
"epoch": 3.93,
"learning_rate": 5.6699999999999996e-05,
"loss": 0.5992,
"step": 570
},
{
"epoch": 4.0,
"learning_rate": 5.77e-05,
"loss": 0.7346,
"step": 580
},
{
"epoch": 4.07,
"learning_rate": 5.87e-05,
"loss": 0.5768,
"step": 590
},
{
"epoch": 4.14,
"learning_rate": 5.97e-05,
"loss": 0.4121,
"step": 600
},
{
"epoch": 4.14,
"eval_loss": 0.5828642845153809,
"eval_runtime": 54.1137,
"eval_samples_per_second": 31.046,
"eval_steps_per_second": 31.046,
"eval_wer": 0.5073392598718214,
"step": 600
},
{
"epoch": 4.21,
"learning_rate": 6.07e-05,
"loss": 0.4987,
"step": 610
},
{
"epoch": 4.28,
"learning_rate": 6.170000000000001e-05,
"loss": 0.65,
"step": 620
},
{
"epoch": 4.34,
"learning_rate": 6.27e-05,
"loss": 0.4376,
"step": 630
},
{
"epoch": 4.41,
"learning_rate": 6.37e-05,
"loss": 0.4473,
"step": 640
},
{
"epoch": 4.48,
"learning_rate": 6.47e-05,
"loss": 0.5292,
"step": 650
},
{
"epoch": 4.55,
"learning_rate": 6.570000000000001e-05,
"loss": 0.541,
"step": 660
},
{
"epoch": 4.62,
"learning_rate": 6.670000000000001e-05,
"loss": 0.3965,
"step": 670
},
{
"epoch": 4.69,
"learning_rate": 6.77e-05,
"loss": 0.4583,
"step": 680
},
{
"epoch": 4.76,
"learning_rate": 6.87e-05,
"loss": 0.6338,
"step": 690
},
{
"epoch": 4.83,
"learning_rate": 6.97e-05,
"loss": 0.3981,
"step": 700
},
{
"epoch": 4.83,
"eval_loss": 0.5153024792671204,
"eval_runtime": 54.0248,
"eval_samples_per_second": 31.097,
"eval_steps_per_second": 31.097,
"eval_wer": 0.44614430432086005,
"step": 700
},
{
"epoch": 4.9,
"learning_rate": 7.07e-05,
"loss": 0.3658,
"step": 710
},
{
"epoch": 4.97,
"learning_rate": 7.17e-05,
"loss": 0.4519,
"step": 720
},
{
"epoch": 5.03,
"learning_rate": 7.27e-05,
"loss": 0.5366,
"step": 730
},
{
"epoch": 5.1,
"learning_rate": 7.37e-05,
"loss": 0.2945,
"step": 740
},
{
"epoch": 5.17,
"learning_rate": 7.47e-05,
"loss": 0.3191,
"step": 750
},
{
"epoch": 5.24,
"learning_rate": 7.570000000000001e-05,
"loss": 0.455,
"step": 760
},
{
"epoch": 5.31,
"learning_rate": 7.670000000000001e-05,
"loss": 0.3723,
"step": 770
},
{
"epoch": 5.38,
"learning_rate": 7.77e-05,
"loss": 0.3053,
"step": 780
},
{
"epoch": 5.45,
"learning_rate": 7.87e-05,
"loss": 0.3479,
"step": 790
},
{
"epoch": 5.52,
"learning_rate": 7.970000000000001e-05,
"loss": 0.5038,
"step": 800
},
{
"epoch": 5.52,
"eval_loss": 0.49075645208358765,
"eval_runtime": 54.3008,
"eval_samples_per_second": 30.939,
"eval_steps_per_second": 30.939,
"eval_wer": 0.415133347115981,
"step": 800
},
{
"epoch": 5.59,
"learning_rate": 8.070000000000001e-05,
"loss": 0.3196,
"step": 810
},
{
"epoch": 5.66,
"learning_rate": 8.17e-05,
"loss": 0.2809,
"step": 820
},
{
"epoch": 5.72,
"learning_rate": 8.27e-05,
"loss": 0.3847,
"step": 830
},
{
"epoch": 5.79,
"learning_rate": 8.37e-05,
"loss": 0.4481,
"step": 840
},
{
"epoch": 5.86,
"learning_rate": 8.47e-05,
"loss": 0.2462,
"step": 850
},
{
"epoch": 5.93,
"learning_rate": 8.57e-05,
"loss": 0.3191,
"step": 860
},
{
"epoch": 6.0,
"learning_rate": 8.67e-05,
"loss": 0.4654,
"step": 870
},
{
"epoch": 6.07,
"learning_rate": 8.77e-05,
"loss": 0.2452,
"step": 880
},
{
"epoch": 6.14,
"learning_rate": 8.87e-05,
"loss": 0.2269,
"step": 890
},
{
"epoch": 6.21,
"learning_rate": 8.970000000000001e-05,
"loss": 0.2899,
"step": 900
},
{
"epoch": 6.21,
"eval_loss": 0.5122061371803284,
"eval_runtime": 54.0537,
"eval_samples_per_second": 31.08,
"eval_steps_per_second": 31.08,
"eval_wer": 0.41106746606023015,
"step": 900
},
{
"epoch": 6.28,
"learning_rate": 9.070000000000001e-05,
"loss": 0.4566,
"step": 910
},
{
"epoch": 6.34,
"learning_rate": 9.17e-05,
"loss": 0.2438,
"step": 920
},
{
"epoch": 6.41,
"learning_rate": 9.27e-05,
"loss": 0.2638,
"step": 930
},
{
"epoch": 6.48,
"learning_rate": 9.370000000000001e-05,
"loss": 0.3491,
"step": 940
},
{
"epoch": 6.55,
"learning_rate": 9.47e-05,
"loss": 0.3234,
"step": 950
},
{
"epoch": 6.62,
"learning_rate": 9.57e-05,
"loss": 0.2101,
"step": 960
},
{
"epoch": 6.69,
"learning_rate": 9.67e-05,
"loss": 0.2624,
"step": 970
},
{
"epoch": 6.76,
"learning_rate": 9.77e-05,
"loss": 0.4167,
"step": 980
},
{
"epoch": 6.83,
"learning_rate": 9.87e-05,
"loss": 0.2703,
"step": 990
},
{
"epoch": 6.9,
"learning_rate": 9.970000000000001e-05,
"loss": 0.2198,
"step": 1000
},
{
"epoch": 6.9,
"eval_loss": 0.4908137619495392,
"eval_runtime": 54.1443,
"eval_samples_per_second": 31.028,
"eval_steps_per_second": 31.028,
"eval_wer": 0.3802632485700503,
"step": 1000
},
{
"epoch": 6.97,
"learning_rate": 9.963157894736843e-05,
"loss": 0.3024,
"step": 1010
},
{
"epoch": 7.03,
"learning_rate": 9.910526315789475e-05,
"loss": 0.3396,
"step": 1020
},
{
"epoch": 7.1,
"learning_rate": 9.857894736842106e-05,
"loss": 0.1892,
"step": 1030
},
{
"epoch": 7.17,
"learning_rate": 9.805263157894737e-05,
"loss": 0.2185,
"step": 1040
},
{
"epoch": 7.24,
"learning_rate": 9.752631578947369e-05,
"loss": 0.3139,
"step": 1050
},
{
"epoch": 7.31,
"learning_rate": 9.7e-05,
"loss": 0.2546,
"step": 1060
},
{
"epoch": 7.38,
"learning_rate": 9.647368421052631e-05,
"loss": 0.1932,
"step": 1070
},
{
"epoch": 7.45,
"learning_rate": 9.594736842105264e-05,
"loss": 0.238,
"step": 1080
},
{
"epoch": 7.52,
"learning_rate": 9.542105263157895e-05,
"loss": 0.3596,
"step": 1090
},
{
"epoch": 7.59,
"learning_rate": 9.489473684210527e-05,
"loss": 0.2129,
"step": 1100
},
{
"epoch": 7.59,
"eval_loss": 0.4668470621109009,
"eval_runtime": 54.8544,
"eval_samples_per_second": 30.627,
"eval_steps_per_second": 30.627,
"eval_wer": 0.37888498380538904,
"step": 1100
},
{
"epoch": 7.66,
"learning_rate": 9.436842105263158e-05,
"loss": 0.2068,
"step": 1110
},
{
"epoch": 7.72,
"learning_rate": 9.384210526315789e-05,
"loss": 0.2542,
"step": 1120
},
{
"epoch": 7.79,
"learning_rate": 9.331578947368422e-05,
"loss": 0.2859,
"step": 1130
},
{
"epoch": 7.86,
"learning_rate": 9.278947368421053e-05,
"loss": 0.1856,
"step": 1140
},
{
"epoch": 7.93,
"learning_rate": 9.226315789473686e-05,
"loss": 0.2141,
"step": 1150
},
{
"epoch": 8.0,
"learning_rate": 9.173684210526317e-05,
"loss": 0.3368,
"step": 1160
},
{
"epoch": 8.07,
"learning_rate": 9.121052631578948e-05,
"loss": 0.1858,
"step": 1170
},
{
"epoch": 8.14,
"learning_rate": 9.06842105263158e-05,
"loss": 0.168,
"step": 1180
},
{
"epoch": 8.21,
"learning_rate": 9.015789473684211e-05,
"loss": 0.1934,
"step": 1190
},
{
"epoch": 8.28,
"learning_rate": 8.963157894736842e-05,
"loss": 0.3007,
"step": 1200
},
{
"epoch": 8.28,
"eval_loss": 0.4788149893283844,
"eval_runtime": 54.1945,
"eval_samples_per_second": 30.999,
"eval_steps_per_second": 30.999,
"eval_wer": 0.35621252842671075,
"step": 1200
},
{
"epoch": 8.34,
"learning_rate": 8.910526315789474e-05,
"loss": 0.1603,
"step": 1210
},
{
"epoch": 8.41,
"learning_rate": 8.857894736842106e-05,
"loss": 0.1688,
"step": 1220
},
{
"epoch": 8.48,
"learning_rate": 8.805263157894737e-05,
"loss": 0.2502,
"step": 1230
},
{
"epoch": 8.55,
"learning_rate": 8.752631578947369e-05,
"loss": 0.264,
"step": 1240
},
{
"epoch": 8.62,
"learning_rate": 8.7e-05,
"loss": 0.1541,
"step": 1250
},
{
"epoch": 8.69,
"learning_rate": 8.647368421052631e-05,
"loss": 0.2043,
"step": 1260
},
{
"epoch": 8.76,
"learning_rate": 8.594736842105263e-05,
"loss": 0.3179,
"step": 1270
},
{
"epoch": 8.83,
"learning_rate": 8.542105263157894e-05,
"loss": 0.1745,
"step": 1280
},
{
"epoch": 8.9,
"learning_rate": 8.489473684210527e-05,
"loss": 0.1525,
"step": 1290
},
{
"epoch": 8.97,
"learning_rate": 8.436842105263158e-05,
"loss": 0.2264,
"step": 1300
},
{
"epoch": 8.97,
"eval_loss": 0.5112716555595398,
"eval_runtime": 54.21,
"eval_samples_per_second": 30.991,
"eval_steps_per_second": 30.991,
"eval_wer": 0.36351733167941563,
"step": 1300
},
{
"epoch": 9.03,
"learning_rate": 8.38421052631579e-05,
"loss": 0.2444,
"step": 1310
},
{
"epoch": 9.1,
"learning_rate": 8.331578947368422e-05,
"loss": 0.1439,
"step": 1320
},
{
"epoch": 9.17,
"learning_rate": 8.278947368421053e-05,
"loss": 0.1469,
"step": 1330
},
{
"epoch": 9.24,
"learning_rate": 8.226315789473684e-05,
"loss": 0.2434,
"step": 1340
},
{
"epoch": 9.31,
"learning_rate": 8.173684210526317e-05,
"loss": 0.2036,
"step": 1350
},
{
"epoch": 9.38,
"learning_rate": 8.121052631578948e-05,
"loss": 0.1351,
"step": 1360
},
{
"epoch": 9.45,
"learning_rate": 8.06842105263158e-05,
"loss": 0.1851,
"step": 1370
},
{
"epoch": 9.52,
"learning_rate": 8.015789473684211e-05,
"loss": 0.2179,
"step": 1380
},
{
"epoch": 9.59,
"learning_rate": 7.963157894736842e-05,
"loss": 0.1332,
"step": 1390
},
{
"epoch": 9.66,
"learning_rate": 7.910526315789474e-05,
"loss": 0.1536,
"step": 1400
},
{
"epoch": 9.66,
"eval_loss": 0.49498996138572693,
"eval_runtime": 54.0568,
"eval_samples_per_second": 31.078,
"eval_steps_per_second": 31.078,
"eval_wer": 0.3440837984976914,
"step": 1400
},
{
"epoch": 9.72,
"learning_rate": 7.857894736842105e-05,
"loss": 0.2393,
"step": 1410
},
{
"epoch": 9.79,
"learning_rate": 7.805263157894738e-05,
"loss": 0.21,
"step": 1420
},
{
"epoch": 9.86,
"learning_rate": 7.752631578947369e-05,
"loss": 0.1434,
"step": 1430
},
{
"epoch": 9.93,
"learning_rate": 7.7e-05,
"loss": 0.1483,
"step": 1440
},
{
"epoch": 10.0,
"learning_rate": 7.647368421052631e-05,
"loss": 0.2758,
"step": 1450
},
{
"epoch": 10.07,
"learning_rate": 7.594736842105263e-05,
"loss": 0.1253,
"step": 1460
},
{
"epoch": 10.14,
"learning_rate": 7.542105263157895e-05,
"loss": 0.1208,
"step": 1470
},
{
"epoch": 10.21,
"learning_rate": 7.489473684210527e-05,
"loss": 0.1522,
"step": 1480
},
{
"epoch": 10.28,
"learning_rate": 7.43684210526316e-05,
"loss": 0.2341,
"step": 1490
},
{
"epoch": 10.34,
"learning_rate": 7.38421052631579e-05,
"loss": 0.1206,
"step": 1500
},
{
"epoch": 10.34,
"eval_loss": 0.5062456130981445,
"eval_runtime": 54.4491,
"eval_samples_per_second": 30.855,
"eval_steps_per_second": 30.855,
"eval_wer": 0.34208531458893254,
"step": 1500
},
{
"epoch": 10.41,
"learning_rate": 7.331578947368422e-05,
"loss": 0.1347,
"step": 1510
},
{
"epoch": 10.48,
"learning_rate": 7.278947368421053e-05,
"loss": 0.1866,
"step": 1520
},
{
"epoch": 10.55,
"learning_rate": 7.226315789473685e-05,
"loss": 0.1933,
"step": 1530
},
{
"epoch": 10.62,
"learning_rate": 7.173684210526316e-05,
"loss": 0.1207,
"step": 1540
},
{
"epoch": 10.69,
"learning_rate": 7.121052631578947e-05,
"loss": 0.1598,
"step": 1550
},
{
"epoch": 10.76,
"learning_rate": 7.06842105263158e-05,
"loss": 0.2104,
"step": 1560
},
{
"epoch": 10.83,
"learning_rate": 7.015789473684211e-05,
"loss": 0.1259,
"step": 1570
},
{
"epoch": 10.9,
"learning_rate": 6.963157894736842e-05,
"loss": 0.1293,
"step": 1580
},
{
"epoch": 10.97,
"learning_rate": 6.910526315789474e-05,
"loss": 0.1797,
"step": 1590
},
{
"epoch": 11.03,
"learning_rate": 6.857894736842105e-05,
"loss": 0.2021,
"step": 1600
},
{
"epoch": 11.03,
"eval_loss": 0.48995423316955566,
"eval_runtime": 54.3979,
"eval_samples_per_second": 30.884,
"eval_steps_per_second": 30.884,
"eval_wer": 0.32830266694231963,
"step": 1600
},
{
"epoch": 11.1,
"learning_rate": 6.805263157894736e-05,
"loss": 0.0889,
"step": 1610
},
{
"epoch": 11.17,
"learning_rate": 6.752631578947368e-05,
"loss": 0.1165,
"step": 1620
},
{
"epoch": 11.24,
"learning_rate": 6.7e-05,
"loss": 0.1852,
"step": 1630
},
{
"epoch": 11.31,
"learning_rate": 6.647368421052632e-05,
"loss": 0.1533,
"step": 1640
},
{
"epoch": 11.38,
"learning_rate": 6.594736842105264e-05,
"loss": 0.1174,
"step": 1650
},
{
"epoch": 11.45,
"learning_rate": 6.542105263157895e-05,
"loss": 0.1412,
"step": 1660
},
{
"epoch": 11.52,
"learning_rate": 6.489473684210527e-05,
"loss": 0.238,
"step": 1670
},
{
"epoch": 11.59,
"learning_rate": 6.436842105263158e-05,
"loss": 0.1124,
"step": 1680
},
{
"epoch": 11.66,
"learning_rate": 6.384210526315791e-05,
"loss": 0.1206,
"step": 1690
},
{
"epoch": 11.72,
"learning_rate": 6.331578947368422e-05,
"loss": 0.1458,
"step": 1700
},
{
"epoch": 11.72,
"eval_loss": 0.5019029974937439,
"eval_runtime": 54.2675,
"eval_samples_per_second": 30.958,
"eval_steps_per_second": 30.958,
"eval_wer": 0.3307146302804769,
"step": 1700
},
{
"epoch": 11.79,
"learning_rate": 6.278947368421053e-05,
"loss": 0.1878,
"step": 1710
},
{
"epoch": 11.86,
"learning_rate": 6.226315789473685e-05,
"loss": 0.1076,
"step": 1720
},
{
"epoch": 11.93,
"learning_rate": 6.173684210526316e-05,
"loss": 0.1281,
"step": 1730
},
{
"epoch": 12.0,
"learning_rate": 6.121052631578947e-05,
"loss": 0.2262,
"step": 1740
},
{
"epoch": 12.07,
"learning_rate": 6.0684210526315785e-05,
"loss": 0.1107,
"step": 1750
},
{
"epoch": 12.14,
"learning_rate": 6.015789473684211e-05,
"loss": 0.1029,
"step": 1760
},
{
"epoch": 12.21,
"learning_rate": 5.9631578947368425e-05,
"loss": 0.1331,
"step": 1770
},
{
"epoch": 12.28,
"learning_rate": 5.9105263157894744e-05,
"loss": 0.173,
"step": 1780
},
{
"epoch": 12.34,
"learning_rate": 5.857894736842106e-05,
"loss": 0.0987,
"step": 1790
},
{
"epoch": 12.41,
"learning_rate": 5.805263157894737e-05,
"loss": 0.1151,
"step": 1800
},
{
"epoch": 12.41,
"eval_loss": 0.49889835715293884,
"eval_runtime": 54.6479,
"eval_samples_per_second": 30.742,
"eval_steps_per_second": 30.742,
"eval_wer": 0.3269933154158914,
"step": 1800
},
{
"epoch": 12.48,
"learning_rate": 5.752631578947368e-05,
"loss": 0.1754,
"step": 1810
},
{
"epoch": 12.55,
"learning_rate": 5.6999999999999996e-05,
"loss": 0.1323,
"step": 1820
},
{
"epoch": 12.62,
"learning_rate": 5.647368421052632e-05,
"loss": 0.0976,
"step": 1830
},
{
"epoch": 12.69,
"learning_rate": 5.5947368421052636e-05,
"loss": 0.1278,
"step": 1840
},
{
"epoch": 12.76,
"learning_rate": 5.542105263157895e-05,
"loss": 0.1764,
"step": 1850
},
{
"epoch": 12.83,
"learning_rate": 5.489473684210527e-05,
"loss": 0.0999,
"step": 1860
},
{
"epoch": 12.9,
"learning_rate": 5.436842105263158e-05,
"loss": 0.0932,
"step": 1870
},
{
"epoch": 12.97,
"learning_rate": 5.3842105263157895e-05,
"loss": 0.1429,
"step": 1880
},
{
"epoch": 13.03,
"learning_rate": 5.331578947368421e-05,
"loss": 0.1598,
"step": 1890
},
{
"epoch": 13.1,
"learning_rate": 5.2789473684210534e-05,
"loss": 0.0985,
"step": 1900
},
{
"epoch": 13.1,
"eval_loss": 0.4924699366092682,
"eval_runtime": 54.4191,
"eval_samples_per_second": 30.872,
"eval_steps_per_second": 30.872,
"eval_wer": 0.31734546206326236,
"step": 1900
},
{
"epoch": 13.17,
"learning_rate": 5.226315789473685e-05,
"loss": 0.1226,
"step": 1910
},
{
"epoch": 13.24,
"learning_rate": 5.173684210526316e-05,
"loss": 0.1576,
"step": 1920
},
{
"epoch": 13.31,
"learning_rate": 5.121052631578947e-05,
"loss": 0.1305,
"step": 1930
},
{
"epoch": 13.38,
"learning_rate": 5.068421052631579e-05,
"loss": 0.0873,
"step": 1940
},
{
"epoch": 13.45,
"learning_rate": 5.0157894736842106e-05,
"loss": 0.1148,
"step": 1950
},
{
"epoch": 13.52,
"learning_rate": 4.9631578947368426e-05,
"loss": 0.1737,
"step": 1960
},
{
"epoch": 13.59,
"learning_rate": 4.910526315789474e-05,
"loss": 0.0993,
"step": 1970
},
{
"epoch": 13.66,
"learning_rate": 4.857894736842106e-05,
"loss": 0.1166,
"step": 1980
},
{
"epoch": 13.72,
"learning_rate": 4.805263157894737e-05,
"loss": 0.1484,
"step": 1990
},
{
"epoch": 13.79,
"learning_rate": 4.7526315789473684e-05,
"loss": 0.1412,
"step": 2000
},
{
"epoch": 13.79,
"eval_loss": 0.4868471920490265,
"eval_runtime": 54.2379,
"eval_samples_per_second": 30.975,
"eval_steps_per_second": 30.975,
"eval_wer": 0.3124526221487148,
"step": 2000
},
{
"epoch": 13.86,
"learning_rate": 4.7e-05,
"loss": 0.0891,
"step": 2010
},
{
"epoch": 13.93,
"learning_rate": 4.647368421052632e-05,
"loss": 0.096,
"step": 2020
},
{
"epoch": 14.0,
"learning_rate": 4.594736842105264e-05,
"loss": 0.1873,
"step": 2030
},
{
"epoch": 14.07,
"learning_rate": 4.542105263157895e-05,
"loss": 0.1013,
"step": 2040
},
{
"epoch": 14.14,
"learning_rate": 4.489473684210527e-05,
"loss": 0.1011,
"step": 2050
},
{
"epoch": 14.21,
"learning_rate": 4.436842105263158e-05,
"loss": 0.1159,
"step": 2060
},
{
"epoch": 14.28,
"learning_rate": 4.3842105263157895e-05,
"loss": 0.1376,
"step": 2070
},
{
"epoch": 14.34,
"learning_rate": 4.3315789473684215e-05,
"loss": 0.0781,
"step": 2080
},
{
"epoch": 14.41,
"learning_rate": 4.278947368421053e-05,
"loss": 0.1111,
"step": 2090
},
{
"epoch": 14.48,
"learning_rate": 4.226315789473684e-05,
"loss": 0.1579,
"step": 2100
},
{
"epoch": 14.48,
"eval_loss": 0.49825519323349,
"eval_runtime": 54.3764,
"eval_samples_per_second": 30.896,
"eval_steps_per_second": 30.896,
"eval_wer": 0.3147267590104059,
"step": 2100
},
{
"epoch": 14.55,
"learning_rate": 4.1736842105263154e-05,
"loss": 0.1385,
"step": 2110
},
{
"epoch": 14.62,
"learning_rate": 4.1210526315789474e-05,
"loss": 0.0803,
"step": 2120
},
{
"epoch": 14.69,
"learning_rate": 4.0684210526315794e-05,
"loss": 0.1259,
"step": 2130
},
{
"epoch": 14.76,
"learning_rate": 4.015789473684211e-05,
"loss": 0.1478,
"step": 2140
},
{
"epoch": 14.83,
"learning_rate": 3.9631578947368426e-05,
"loss": 0.0832,
"step": 2150
},
{
"epoch": 14.9,
"learning_rate": 3.910526315789474e-05,
"loss": 0.0781,
"step": 2160
},
{
"epoch": 14.97,
"learning_rate": 3.857894736842105e-05,
"loss": 0.1148,
"step": 2170
},
{
"epoch": 15.03,
"learning_rate": 3.8052631578947365e-05,
"loss": 0.1716,
"step": 2180
},
{
"epoch": 15.1,
"learning_rate": 3.7526315789473685e-05,
"loss": 0.0828,
"step": 2190
},
{
"epoch": 15.17,
"learning_rate": 3.7e-05,
"loss": 0.1043,
"step": 2200
},
{
"epoch": 15.17,
"eval_loss": 0.4913721978664398,
"eval_runtime": 54.4035,
"eval_samples_per_second": 30.88,
"eval_steps_per_second": 30.88,
"eval_wer": 0.3091447867135277,
"step": 2200
},
{
"epoch": 15.24,
"learning_rate": 3.647368421052632e-05,
"loss": 0.1304,
"step": 2210
},
{
"epoch": 15.31,
"learning_rate": 3.594736842105264e-05,
"loss": 0.109,
"step": 2220
},
{
"epoch": 15.38,
"learning_rate": 3.542105263157895e-05,
"loss": 0.0837,
"step": 2230
},
{
"epoch": 15.45,
"learning_rate": 3.4894736842105264e-05,
"loss": 0.1096,
"step": 2240
},
{
"epoch": 15.52,
"learning_rate": 3.436842105263158e-05,
"loss": 0.1639,
"step": 2250
},
{
"epoch": 15.59,
"learning_rate": 3.3842105263157896e-05,
"loss": 0.0872,
"step": 2260
},
{
"epoch": 15.66,
"learning_rate": 3.331578947368421e-05,
"loss": 0.0837,
"step": 2270
},
{
"epoch": 15.72,
"learning_rate": 3.278947368421052e-05,
"loss": 0.1221,
"step": 2280
},
{
"epoch": 15.79,
"learning_rate": 3.226315789473684e-05,
"loss": 0.1432,
"step": 2290
},
{
"epoch": 15.86,
"learning_rate": 3.173684210526316e-05,
"loss": 0.0773,
"step": 2300
},
{
"epoch": 15.86,
"eval_loss": 0.48580366373062134,
"eval_runtime": 54.3133,
"eval_samples_per_second": 30.932,
"eval_steps_per_second": 30.932,
"eval_wer": 0.3102473985252567,
"step": 2300
},
{
"epoch": 15.93,
"learning_rate": 3.1210526315789475e-05,
"loss": 0.0854,
"step": 2310
},
{
"epoch": 16.0,
"learning_rate": 3.0684210526315795e-05,
"loss": 0.1744,
"step": 2320
},
{
"epoch": 16.07,
"learning_rate": 3.0157894736842108e-05,
"loss": 0.0966,
"step": 2330
},
{
"epoch": 16.14,
"learning_rate": 2.963157894736842e-05,
"loss": 0.0964,
"step": 2340
},
{
"epoch": 16.21,
"learning_rate": 2.910526315789474e-05,
"loss": 0.1164,
"step": 2350
},
{
"epoch": 16.28,
"learning_rate": 2.8578947368421057e-05,
"loss": 0.1278,
"step": 2360
},
{
"epoch": 16.34,
"learning_rate": 2.805263157894737e-05,
"loss": 0.0853,
"step": 2370
},
{
"epoch": 16.41,
"learning_rate": 2.7526315789473683e-05,
"loss": 0.1002,
"step": 2380
},
{
"epoch": 16.48,
"learning_rate": 2.7000000000000002e-05,
"loss": 0.0967,
"step": 2390
},
{
"epoch": 16.55,
"learning_rate": 2.647368421052632e-05,
"loss": 0.1327,
"step": 2400
},
{
"epoch": 16.55,
"eval_loss": 0.508406400680542,
"eval_runtime": 54.6615,
"eval_samples_per_second": 30.735,
"eval_steps_per_second": 30.735,
"eval_wer": 0.3063882571842051,
"step": 2400
},
{
"epoch": 16.62,
"learning_rate": 2.5947368421052632e-05,
"loss": 0.0717,
"step": 2410
},
{
"epoch": 16.69,
"learning_rate": 2.542105263157895e-05,
"loss": 0.1023,
"step": 2420
},
{
"epoch": 16.76,
"learning_rate": 2.4894736842105264e-05,
"loss": 0.1276,
"step": 2430
},
{
"epoch": 16.83,
"learning_rate": 2.436842105263158e-05,
"loss": 0.0921,
"step": 2440
},
{
"epoch": 16.9,
"learning_rate": 2.3842105263157897e-05,
"loss": 0.0772,
"step": 2450
},
{
"epoch": 16.97,
"learning_rate": 2.331578947368421e-05,
"loss": 0.1152,
"step": 2460
},
{
"epoch": 17.03,
"learning_rate": 2.2789473684210527e-05,
"loss": 0.1465,
"step": 2470
},
{
"epoch": 17.1,
"learning_rate": 2.2263157894736843e-05,
"loss": 0.0677,
"step": 2480
},
{
"epoch": 17.17,
"learning_rate": 2.173684210526316e-05,
"loss": 0.0877,
"step": 2490
},
{
"epoch": 17.24,
"learning_rate": 2.1210526315789476e-05,
"loss": 0.1281,
"step": 2500
},
{
"epoch": 17.24,
"eval_loss": 0.5016999840736389,
"eval_runtime": 54.5027,
"eval_samples_per_second": 30.824,
"eval_steps_per_second": 30.824,
"eval_wer": 0.3024602026049204,
"step": 2500
},
{
"epoch": 17.31,
"learning_rate": 2.068421052631579e-05,
"loss": 0.0974,
"step": 2510
},
{
"epoch": 17.38,
"learning_rate": 2.0157894736842105e-05,
"loss": 0.0766,
"step": 2520
},
{
"epoch": 17.45,
"learning_rate": 1.963157894736842e-05,
"loss": 0.0891,
"step": 2530
},
{
"epoch": 17.52,
"learning_rate": 1.9105263157894738e-05,
"loss": 0.1066,
"step": 2540
},
{
"epoch": 17.59,
"learning_rate": 1.8578947368421054e-05,
"loss": 0.0732,
"step": 2550
},
{
"epoch": 17.66,
"learning_rate": 1.8052631578947367e-05,
"loss": 0.0947,
"step": 2560
},
{
"epoch": 17.72,
"learning_rate": 1.7526315789473683e-05,
"loss": 0.1138,
"step": 2570
},
{
"epoch": 17.79,
"learning_rate": 1.7000000000000003e-05,
"loss": 0.1179,
"step": 2580
},
{
"epoch": 17.86,
"learning_rate": 1.6473684210526316e-05,
"loss": 0.0652,
"step": 2590
},
{
"epoch": 17.93,
"learning_rate": 1.5947368421052633e-05,
"loss": 0.0845,
"step": 2600
},
{
"epoch": 17.93,
"eval_loss": 0.5000839829444885,
"eval_runtime": 54.1449,
"eval_samples_per_second": 31.028,
"eval_steps_per_second": 31.028,
"eval_wer": 0.30121976431672526,
"step": 2600
},
{
"epoch": 18.0,
"learning_rate": 1.5421052631578946e-05,
"loss": 0.138,
"step": 2610
},
{
"epoch": 18.07,
"learning_rate": 1.4894736842105264e-05,
"loss": 0.082,
"step": 2620
},
{
"epoch": 18.14,
"learning_rate": 1.4368421052631582e-05,
"loss": 0.0715,
"step": 2630
},
{
"epoch": 18.21,
"learning_rate": 1.3842105263157895e-05,
"loss": 0.1127,
"step": 2640
},
{
"epoch": 18.28,
"learning_rate": 1.3315789473684213e-05,
"loss": 0.1289,
"step": 2650
},
{
"epoch": 18.34,
"learning_rate": 1.2789473684210526e-05,
"loss": 0.0795,
"step": 2660
},
{
"epoch": 18.41,
"learning_rate": 1.2263157894736844e-05,
"loss": 0.0771,
"step": 2670
},
{
"epoch": 18.48,
"learning_rate": 1.1736842105263158e-05,
"loss": 0.135,
"step": 2680
},
{
"epoch": 18.55,
"learning_rate": 1.1210526315789475e-05,
"loss": 0.0968,
"step": 2690
},
{
"epoch": 18.62,
"learning_rate": 1.068421052631579e-05,
"loss": 0.0717,
"step": 2700
},
{
"epoch": 18.62,
"eval_loss": 0.4893862009048462,
"eval_runtime": 54.1846,
"eval_samples_per_second": 31.005,
"eval_steps_per_second": 31.005,
"eval_wer": 0.30039280545792846,
"step": 2700
},
{
"epoch": 18.69,
"learning_rate": 1.0157894736842106e-05,
"loss": 0.0902,
"step": 2710
},
{
"epoch": 18.76,
"learning_rate": 9.631578947368422e-06,
"loss": 0.1204,
"step": 2720
},
{
"epoch": 18.83,
"learning_rate": 9.105263157894737e-06,
"loss": 0.0536,
"step": 2730
},
{
"epoch": 18.9,
"learning_rate": 8.578947368421053e-06,
"loss": 0.0785,
"step": 2740
},
{
"epoch": 18.97,
"learning_rate": 8.052631578947368e-06,
"loss": 0.0987,
"step": 2750
},
{
"epoch": 19.03,
"learning_rate": 7.526315789473684e-06,
"loss": 0.117,
"step": 2760
},
{
"epoch": 19.1,
"learning_rate": 7.000000000000001e-06,
"loss": 0.0767,
"step": 2770
},
{
"epoch": 19.17,
"learning_rate": 6.473684210526316e-06,
"loss": 0.0718,
"step": 2780
},
{
"epoch": 19.24,
"learning_rate": 5.947368421052632e-06,
"loss": 0.1384,
"step": 2790
},
{
"epoch": 19.31,
"learning_rate": 5.421052631578947e-06,
"loss": 0.0835,
"step": 2800
},
{
"epoch": 19.31,
"eval_loss": 0.4963100254535675,
"eval_runtime": 54.5341,
"eval_samples_per_second": 30.806,
"eval_steps_per_second": 30.806,
"eval_wer": 0.29984149955206396,
"step": 2800
},
{
"epoch": 19.38,
"learning_rate": 4.894736842105263e-06,
"loss": 0.0742,
"step": 2810
},
{
"epoch": 19.45,
"learning_rate": 4.368421052631579e-06,
"loss": 0.1011,
"step": 2820
},
{
"epoch": 19.52,
"learning_rate": 3.842105263157895e-06,
"loss": 0.1401,
"step": 2830
},
{
"epoch": 19.59,
"learning_rate": 3.315789473684211e-06,
"loss": 0.0767,
"step": 2840
},
{
"epoch": 19.66,
"learning_rate": 2.7894736842105266e-06,
"loss": 0.078,
"step": 2850
},
{
"epoch": 19.72,
"learning_rate": 2.2631578947368426e-06,
"loss": 0.0925,
"step": 2860
},
{
"epoch": 19.79,
"learning_rate": 1.7368421052631579e-06,
"loss": 0.1296,
"step": 2870
},
{
"epoch": 19.86,
"learning_rate": 1.2105263157894738e-06,
"loss": 0.07,
"step": 2880
},
{
"epoch": 19.93,
"learning_rate": 6.842105263157896e-07,
"loss": 0.0859,
"step": 2890
},
{
"epoch": 20.0,
"learning_rate": 1.5789473684210527e-07,
"loss": 0.1181,
"step": 2900
},
{
"epoch": 20.0,
"eval_loss": 0.49261119961738586,
"eval_runtime": 53.4045,
"eval_samples_per_second": 31.458,
"eval_steps_per_second": 31.458,
"eval_wer": 0.29880780097856796,
"step": 2900
},
{
"epoch": 20.0,
"step": 2900,
"total_flos": 2.523397902383489e+18,
"train_loss": 0.6981826857246202,
"train_runtime": 3982.7046,
"train_samples_per_second": 23.2,
"train_steps_per_second": 0.728
}
],
"max_steps": 2900,
"num_train_epochs": 20,
"total_flos": 2.523397902383489e+18,
"trial_name": null,
"trial_params": null
}