xls-r-asr_af-run8
This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on the asr_af dataset. It achieves the following results:
- Wer (Validation): 37.00%
- Wer (Test): 38.77%
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0003
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- 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 (Train) |
|---|---|---|---|---|
| 4.1505 | 2.35 | 400 | 1.0968 | 0.7889 |
| 0.6096 | 4.69 | 800 | 0.5411 | 0.4601 |
| 0.2559 | 7.04 | 1200 | 0.4883 | 0.4014 |
| 0.1464 | 9.38 | 1600 | 0.4516 | 0.3704 |
| 0.1079 | 11.73 | 2000 | 0.4880 | 0.3563 |
| 0.0808 | 14.08 | 2400 | 0.4873 | 0.3526 |
Framework versions
- Transformers 4.28.0
- Pytorch 2.0.1+cu117
- Datasets 2.14.4
- Tokenizers 0.13.3
- Downloads last month
- 7
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.