wav2vec2-large-xlsr-turkish-demo-colab
This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the common_voice dataset. It achieves the following results on the evaluation set:
- Loss: 0.4151
- Wer: 0.4822
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: 30
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 5.2487 | 4.26 | 400 | 1.6455 | 1.0778 |
| 0.71 | 8.51 | 800 | 0.4428 | 0.6138 |
| 0.3073 | 12.77 | 1200 | 0.4214 | 0.5517 |
| 0.2136 | 17.02 | 1600 | 0.4345 | 0.5193 |
| 0.1624 | 21.28 | 2000 | 0.4366 | 0.5026 |
| 0.1298 | 25.53 | 2400 | 0.4111 | 0.4949 |
| 0.1174 | 29.79 | 2800 | 0.4151 | 0.4822 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1+cu116
- Datasets 2.10.1
- Tokenizers 0.13.2