<!-- 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-m02-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: 0.0260
- Wer: 0.4968
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 |
---|---|---|---|---|
3.4385 | 0.91 | 500 | 4.0160 | 1.0 |
3.0413 | 1.81 | 1000 | 3.2881 | 1.0 |
3.0011 | 2.72 | 1500 | 3.2401 | 1.0 |
2.8653 | 3.62 | 2000 | 3.0338 | 1.0 |
2.6386 | 4.53 | 2500 | 2.7806 | 1.0492 |
2.5376 | 5.43 | 3000 | 2.5253 | 1.3647 |
2.2722 | 6.34 | 3500 | 2.1425 | 1.3252 |
1.627 | 7.25 | 4000 | 1.4101 | 1.3658 |
1.2689 | 8.15 | 4500 | 0.9284 | 1.2448 |
1.0197 | 9.06 | 5000 | 0.6370 | 1.1254 |
0.8198 | 9.96 | 5500 | 0.4743 | 0.9947 |
0.7357 | 10.87 | 6000 | 0.3423 | 0.8820 |
0.5532 | 11.78 | 6500 | 0.2764 | 0.8203 |
0.5133 | 12.68 | 7000 | 0.2158 | 0.7580 |
0.4943 | 13.59 | 7500 | 0.1872 | 0.7195 |
0.3741 | 14.49 | 8000 | 0.1529 | 0.6762 |
0.3524 | 15.4 | 8500 | 0.1269 | 0.6527 |
0.3086 | 16.3 | 9000 | 0.1049 | 0.6254 |
0.3141 | 17.21 | 9500 | 0.0887 | 0.6012 |
0.2879 | 18.12 | 10000 | 0.0829 | 0.5863 |
0.3141 | 19.02 | 10500 | 0.0660 | 0.5688 |
0.2609 | 19.93 | 11000 | 0.0732 | 0.5591 |
0.2707 | 20.83 | 11500 | 0.0552 | 0.5434 |
0.2307 | 21.74 | 12000 | 0.0524 | 0.5406 |
0.1863 | 22.64 | 12500 | 0.0466 | 0.5281 |
0.2211 | 23.55 | 13000 | 0.0426 | 0.5226 |
0.1827 | 24.46 | 13500 | 0.0365 | 0.5129 |
0.1782 | 25.36 | 14000 | 0.0356 | 0.5099 |
0.1799 | 26.27 | 14500 | 0.0323 | 0.5049 |
0.1481 | 27.17 | 15000 | 0.0300 | 0.5034 |
0.1609 | 28.08 | 15500 | 0.0278 | 0.5030 |
0.1752 | 28.99 | 16000 | 0.0269 | 0.4978 |
0.1541 | 29.89 | 16500 | 0.0260 | 0.4968 |
Framework versions
- Transformers 4.23.1
- Pytorch 1.12.1+cu113
- Datasets 2.0.0
- Tokenizers 0.13.2