<!-- 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-cy
This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the PYTHON/CUSTOM_COMMON_VOICE.PY - CY dataset. It achieves the following results on the evaluation set:
- Loss: 0.5770
- Wer: 0.2776
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: 96
- eval_batch_size: 96
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 192
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 800
- training_steps: 5000
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
5.0439 | 9.76 | 400 | 1.0820 | 0.8189 |
0.3314 | 19.51 | 800 | 0.5394 | 0.4411 |
0.114 | 29.27 | 1200 | 0.5568 | 0.3806 |
0.075 | 39.02 | 1600 | 0.5549 | 0.3513 |
0.0573 | 48.78 | 2000 | 0.5357 | 0.3289 |
0.0454 | 58.54 | 2400 | 0.5619 | 0.3190 |
0.0364 | 68.29 | 2800 | 0.5364 | 0.3157 |
0.0291 | 78.05 | 3200 | 0.5647 | 0.3054 |
0.0237 | 87.8 | 3600 | 0.5450 | 0.2934 |
0.019 | 97.56 | 4000 | 0.5545 | 0.2864 |
0.0157 | 107.32 | 4400 | 0.5574 | 0.2846 |
0.0128 | 117.07 | 4800 | 0.5692 | 0.2792 |
Framework versions
- Transformers 4.33.3
- Pytorch 2.0.1+cu117
- Datasets 2.14.5
- Tokenizers 0.13.3