wav2vec2-common_voice-en-demo-1
This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.8249
- Wer: 0.4713
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: 0.0003
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 3.0
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 3.4837 | 0.4566 | 100 | 3.7538 | 1.0 |
| 2.9754 | 0.9132 | 200 | 3.0914 | 1.0 |
| 2.8257 | 1.3699 | 300 | 2.8550 | 1.0 |
| 0.7154 | 1.8265 | 400 | 1.0927 | 0.6141 |
| 0.5289 | 2.2831 | 500 | 0.9329 | 0.5331 |
| 0.4774 | 2.7397 | 600 | 0.8249 | 0.4713 |
Framework versions
- Transformers 4.43.0.dev0
- Pytorch 2.3.1+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1