metadata
license: apache-2.0
base_model: facebook/hubert-large-ls960-ft
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: hubert_arabic_mdd
results: []
hubert_arabic_mdd
This model is a fine-tuned version of facebook/hubert-large-ls960-ft on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.5404
- Wer: 0.0859
- Per: 0.0671
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: 2
- eval_batch_size: 2
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Per |
|---|---|---|---|---|---|
| 6.3943 | 1.0 | 1637 | 1.2722 | 0.4693 | 0.4456 |
| 0.7962 | 2.0 | 3274 | 0.5990 | 0.1377 | 0.1185 |
| 0.4245 | 3.0 | 4911 | 0.6075 | 0.0899 | 0.0674 |
| 0.2898 | 4.0 | 6548 | 0.5285 | 0.0979 | 0.0738 |
| 0.2262 | 5.0 | 8185 | 0.5600 | 0.0977 | 0.0758 |
| 0.1803 | 6.0 | 9822 | 0.5504 | 0.0808 | 0.0603 |
| 0.1488 | 7.0 | 11459 | 0.5854 | 0.0898 | 0.0700 |
| 0.1267 | 8.0 | 13096 | 0.5438 | 0.0914 | 0.0722 |
| 0.1156 | 9.0 | 14733 | 0.5395 | 0.0866 | 0.0671 |
| 0.0993 | 10.0 | 16370 | 0.5404 | 0.0859 | 0.0671 |
Framework versions
- Transformers 4.35.0
- Pytorch 2.1.0+cu118
- Datasets 2.14.6
- Tokenizers 0.14.1