<!-- 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-xls-r-300m-tonga-test_v2
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: 0.4784
- Wer: 0.3887
- Cer: 0.0940
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: 2
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 40
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|---|---|---|---|---|---|
| 6.6795 | 3.62 | 500 | 2.9371 | 1.0 | 1.0 |
| 2.9201 | 7.25 | 1000 | 2.8427 | 1.0 | 1.0 |
| 2.6938 | 10.87 | 1500 | 1.6109 | 1.0 | 0.3506 |
| 1.1605 | 14.49 | 2000 | 0.6472 | 0.5433 | 0.1229 |
| 0.7598 | 18.12 | 2500 | 0.5543 | 0.4611 | 0.1092 |
| 0.6529 | 21.74 | 3000 | 0.5202 | 0.4289 | 0.1026 |
| 0.5816 | 25.36 | 3500 | 0.4931 | 0.4093 | 0.0981 |
| 0.5476 | 28.98 | 4000 | 0.4916 | 0.4041 | 0.0978 |
| 0.517 | 32.61 | 4500 | 0.4765 | 0.3948 | 0.0951 |
| 0.5025 | 36.23 | 5000 | 0.4812 | 0.3879 | 0.0942 |
| 0.4879 | 39.85 | 5500 | 0.4784 | 0.3887 | 0.0940 |
Framework versions
- Transformers 4.17.0
- Pytorch 1.12.1+cu113
- Datasets 1.18.3
- Tokenizers 0.13.2
