metadata
library_name: transformers
license: mit
base_model: facebook/w2v-bert-2.0
tags:
- generated_from_trainer
metrics:
- wer
language:
- mos
datasets:
- madoss/faso-speech
model-index:
- name: wav2vec-finetuned
results: []
wav2vec-finetuned
This model is a fine-tuned version of facebook/w2v-bert-2.0 on madoss/faso-speech.
It achieves the following results on the evaluation set:
- Loss: 0.6286
- Wer: 0.2902
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 2.9524 | 0.9788 | 300 | 1.1082 | 0.7286 |
| 1.8613 | 1.9560 | 600 | 0.7658 | 0.4447 |
| 1.3791 | 2.9331 | 900 | 0.6838 | 0.3803 |
| 1.0644 | 3.9103 | 1200 | 0.6384 | 0.3221 |
| 0.8971 | 4.8874 | 1500 | 0.6323 | 0.2908 |
| 0.8971 | 5.0 | 1535 | 0.6286 | 0.2902 |
Framework versions
- Transformers 5.9.0
- Pytorch 2.10.0+cu128
- Datasets 4.8.5
- Tokenizers 0.22.2