<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
libri-alpha-0-Temp-1-kl
This model is a fine-tuned version of on the None dataset. It achieves the following results on the evaluation set:
- Loss: 101.1091
- Wer: 0.1503
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 28
- eval_batch_size: 28
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 56
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 30
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
714.4386 | 0.65 | 100 | 206.0585 | 0.1904 |
455.7244 | 1.31 | 200 | 150.6093 | 0.2306 |
397.5524 | 1.96 | 300 | 134.8838 | 0.2301 |
364.2438 | 2.61 | 400 | 125.3695 | 0.2193 |
336.6832 | 3.27 | 500 | 120.0034 | 0.2168 |
337.777 | 3.92 | 600 | 117.2567 | 0.2011 |
315.4012 | 4.58 | 700 | 115.6776 | 0.1879 |
316.353 | 5.23 | 800 | 114.1418 | 0.1826 |
312.6547 | 5.88 | 900 | 112.3635 | 0.1872 |
289.6309 | 6.54 | 1000 | 111.3456 | 0.1741 |
301.9763 | 7.19 | 1100 | 110.7650 | 0.1714 |
289.428 | 7.84 | 1200 | 109.2358 | 0.1709 |
275.0852 | 8.5 | 1300 | 108.4552 | 0.1725 |
297.2818 | 9.15 | 1400 | 108.1707 | 0.1657 |
273.9007 | 9.8 | 1500 | 107.2497 | 0.1616 |
283.6222 | 10.46 | 1600 | 107.5092 | 0.1599 |
279.0072 | 11.11 | 1700 | 106.6930 | 0.1612 |
270.9222 | 11.76 | 1800 | 105.6633 | 0.1635 |
284.9869 | 12.42 | 1900 | 105.7896 | 0.1625 |
271.7268 | 13.07 | 2000 | 105.5718 | 0.1586 |
268.3919 | 13.73 | 2100 | 105.5480 | 0.1591 |
269.6929 | 14.38 | 2200 | 104.9802 | 0.1596 |
260.1387 | 15.03 | 2300 | 104.3509 | 0.1547 |
267.38 | 15.69 | 2400 | 103.6222 | 0.1573 |
264.2958 | 16.34 | 2500 | 103.7570 | 0.1524 |
253.4907 | 16.99 | 2600 | 102.8282 | 0.1580 |
270.6275 | 17.65 | 2700 | 103.1811 | 0.1566 |
257.2584 | 18.3 | 2800 | 102.8419 | 0.1522 |
264.0412 | 18.95 | 2900 | 102.2693 | 0.1551 |
264.1277 | 19.61 | 3000 | 102.4851 | 0.1535 |
250.8255 | 20.26 | 3100 | 101.7220 | 0.1550 |
265.9928 | 20.92 | 3200 | 101.8507 | 0.1541 |
251.8845 | 21.57 | 3300 | 102.0956 | 0.1508 |
256.2606 | 22.22 | 3400 | 101.8346 | 0.1524 |
255.3002 | 22.88 | 3500 | 101.6648 | 0.1495 |
252.5451 | 23.53 | 3600 | 101.2943 | 0.1524 |
256.3488 | 24.18 | 3700 | 101.3040 | 0.1521 |
252.4813 | 24.84 | 3800 | 101.3680 | 0.1502 |
249.9003 | 25.49 | 3900 | 101.1692 | 0.1522 |
254.8601 | 26.14 | 4000 | 101.2368 | 0.1489 |
247.9549 | 26.8 | 4100 | 101.0358 | 0.1516 |
254.5287 | 27.45 | 4200 | 101.1500 | 0.1502 |
255.1176 | 28.1 | 4300 | 101.0062 | 0.1511 |
246.2381 | 28.76 | 4400 | 100.9802 | 0.1505 |
262.6459 | 29.41 | 4500 | 101.1091 | 0.1503 |
Framework versions
- Transformers 4.25.1
- Pytorch 1.12.1
- Datasets 2.7.1
- Tokenizers 0.11.0