<!-- 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-xlsr-53-ft-ccv-btb-cy
This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the CUSTOM_COMMON_VOICE.PY,TECHIAITH/BANC-TRAWSGRIFIADAU-BANGOR - CY dataset. It achieves the following results on the evaluation set:
- Loss: 0.1729
- Wer: 0.1562
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: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 800
- training_steps: 5000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 5.3855 | 0.28 | 400 | 1.3323 | 0.8624 |
| 0.6301 | 0.56 | 800 | 0.4340 | 0.3828 |
| 0.4144 | 0.84 | 1200 | 0.3096 | 0.2935 |
| 0.3277 | 1.12 | 1600 | 0.2646 | 0.2432 |
| 0.2751 | 1.4 | 2000 | 0.2415 | 0.2251 |
| 0.2535 | 1.68 | 2400 | 0.2237 | 0.2094 |
| 0.2409 | 1.96 | 2800 | 0.2092 | 0.1968 |
| 0.1927 | 2.25 | 3200 | 0.1996 | 0.1840 |
| 0.1802 | 2.53 | 3600 | 0.1892 | 0.1762 |
| 0.171 | 2.81 | 4000 | 0.1815 | 0.1670 |
| 0.1521 | 3.09 | 4400 | 0.1789 | 0.1622 |
| 0.1327 | 3.37 | 4800 | 0.1742 | 0.1573 |
Framework versions
- Transformers 4.29.2
- Pytorch 2.0.1+cu117
- Datasets 2.12.0
- Tokenizers 0.13.3
