<!-- 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-burak-new-300-v2-5
This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.2833
- Wer: 0.2168
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: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- 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: 500
- num_epochs: 141
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 6.3785 | 9.8 | 500 | 3.1380 | 1.0 |
| 1.863 | 19.6 | 1000 | 0.3638 | 0.4659 |
| 0.524 | 29.41 | 1500 | 0.2742 | 0.3379 |
| 0.3581 | 39.21 | 2000 | 0.2746 | 0.3049 |
| 0.2783 | 49.02 | 2500 | 0.2559 | 0.2877 |
| 0.2378 | 58.82 | 3000 | 0.2613 | 0.2732 |
| 0.2062 | 68.62 | 3500 | 0.2499 | 0.2602 |
| 0.1849 | 78.43 | 4000 | 0.2809 | 0.2485 |
| 0.1663 | 88.23 | 4500 | 0.2768 | 0.2429 |
| 0.1526 | 98.04 | 5000 | 0.2767 | 0.2319 |
| 0.1434 | 107.84 | 5500 | 0.2886 | 0.2285 |
| 0.1338 | 117.64 | 6000 | 0.2808 | 0.2257 |
| 0.1313 | 127.45 | 6500 | 0.2835 | 0.2106 |
| 0.1281 | 137.25 | 7000 | 0.2833 | 0.2168 |
Framework versions
- Transformers 4.24.0
- Pytorch 1.12.1+cu113
- Datasets 2.6.1
- Tokenizers 0.13.2
