<!-- 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. -->
wav2vec2-large-xlsr-53-torgo-demo-m04-nolm
This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0179
- Wer: 0.4609
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: 0.0001
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- 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 |
|---|---|---|---|---|
| 3.3456 | 0.88 | 500 | 4.3417 | 1.0 |
| 2.9307 | 1.75 | 1000 | 2.9493 | 1.0 |
| 2.741 | 2.63 | 1500 | 2.9803 | 1.0 |
| 2.2291 | 3.5 | 2000 | 1.9442 | 1.1343 |
| 1.3963 | 4.38 | 2500 | 1.0394 | 1.3114 |
| 0.9695 | 5.25 | 3000 | 0.6343 | 1.0388 |
| 0.7907 | 6.13 | 3500 | 0.4243 | 0.9229 |
| 0.6529 | 7.01 | 4000 | 0.3108 | 0.8083 |
| 0.5078 | 7.88 | 4500 | 0.2490 | 0.7440 |
| 0.5039 | 8.76 | 5000 | 0.1810 | 0.7036 |
| 0.4288 | 9.63 | 5500 | 0.1525 | 0.6676 |
| 0.3999 | 10.51 | 6000 | 0.1335 | 0.6371 |
| 0.319 | 11.38 | 6500 | 0.1227 | 0.6016 |
| 0.3104 | 12.26 | 7000 | 0.1064 | 0.5910 |
| 0.3006 | 13.13 | 7500 | 0.0924 | 0.5816 |
| 0.3059 | 14.01 | 8000 | 0.0929 | 0.5604 |
| 0.2689 | 14.89 | 8500 | 0.0692 | 0.5547 |
| 0.2416 | 15.76 | 9000 | 0.0638 | 0.5338 |
| 0.1964 | 16.64 | 9500 | 0.0594 | 0.5233 |
| 0.1868 | 17.51 | 10000 | 0.0535 | 0.5151 |
| 0.1963 | 18.39 | 10500 | 0.0534 | 0.4941 |
| 0.2159 | 19.26 | 11000 | 0.0440 | 0.4890 |
| 0.2059 | 20.14 | 11500 | 0.0432 | 0.4836 |
| 0.1701 | 21.02 | 12000 | 0.0364 | 0.4793 |
| 0.1748 | 21.89 | 12500 | 0.0308 | 0.4763 |
| 0.1673 | 22.77 | 13000 | 0.0266 | 0.4745 |
| 0.1473 | 23.64 | 13500 | 0.0280 | 0.4726 |
| 0.1245 | 24.52 | 14000 | 0.0234 | 0.4700 |
| 0.1396 | 25.39 | 14500 | 0.0243 | 0.4658 |
| 0.1428 | 26.27 | 15000 | 0.0235 | 0.4637 |
| 0.116 | 27.15 | 15500 | 0.0192 | 0.4620 |
| 0.1366 | 28.02 | 16000 | 0.0182 | 0.4603 |
| 0.0974 | 28.9 | 16500 | 0.0186 | 0.4604 |
| 0.1439 | 29.77 | 17000 | 0.0179 | 0.4609 |
Framework versions
- Transformers 4.23.1
- Pytorch 1.12.1+cu113
- Datasets 2.0.0
- Tokenizers 0.13.2
