<!-- 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-large-xls-r-300m-tr
This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on the common_voice dataset. It achieves the following results on the evaluation set:
- Loss: 0.2841
- Wer: 0.2904
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: 3e-05
- train_batch_size: 7
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 14
- 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 |
|---|---|---|---|---|
| 5.0805 | 4.03 | 1000 | 3.0333 | 1.0 |
| 1.5733 | 8.06 | 2000 | 0.5545 | 0.5080 |
| 0.6238 | 12.1 | 3000 | 0.3861 | 0.3977 |
| 0.4535 | 16.13 | 4000 | 0.3253 | 0.3408 |
| 0.3682 | 20.16 | 5000 | 0.3042 | 0.3177 |
| 0.3302 | 24.19 | 6000 | 0.2950 | 0.3015 |
| 0.2985 | 28.23 | 7000 | 0.2841 | 0.2904 |
Framework versions
- Transformers 4.11.3
- Pytorch 1.10.0+cu113
- Datasets 1.18.3
- Tokenizers 0.10.3
