metadata
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: wav2vec2-large-xls-r-300m-Arabic-phoneme
results: []
wav2vec2-large-xls-r-300m-Arabic-phoneme
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: 0.0955
- Per: 0.0852
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: 0.0005
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- 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: 250
- num_epochs: 30
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Per |
|---|---|---|---|---|
| 2.5617 | 1.0 | 102 | 2.2924 | 1.0 |
| 2.2639 | 2.0 | 204 | 2.3079 | 1.0 |
| 2.3203 | 2.99 | 306 | 2.3406 | 1.0 |
| 2.2545 | 4.0 | 409 | 2.2614 | 1.0 |
| 2.1643 | 5.0 | 511 | 1.7664 | 1.0 |
| 1.7206 | 6.0 | 613 | 1.7111 | 1.0 |
| 1.7024 | 6.99 | 715 | 1.7146 | 1.0 |
| 1.719 | 8.0 | 818 | 1.7485 | 1.0 |
| 1.7268 | 9.0 | 920 | 1.7347 | 1.0 |
| 1.7042 | 10.0 | 1022 | 1.7078 | 1.0 |
| 1.6836 | 10.99 | 1124 | 1.6868 | 1.0 |
| 1.6719 | 12.0 | 1227 | 1.6798 | 1.0 |
| 1.6573 | 13.0 | 1329 | 1.6622 | 1.0 |
| 1.6353 | 14.0 | 1431 | 1.6339 | 1.0 |
| 1.6179 | 14.99 | 1533 | 1.5896 | 1.0 |
| 1.5669 | 16.0 | 1636 | 1.5226 | 0.9793 |
| 1.5276 | 17.0 | 1738 | 1.4972 | 0.9533 |
| 1.4857 | 18.0 | 1840 | 1.4075 | 0.9757 |
| 1.4144 | 18.99 | 1942 | 1.3158 | 0.9685 |
| 1.3073 | 20.0 | 2045 | 1.1385 | 0.9692 |
| 1.1651 | 21.0 | 2147 | 0.9220 | 0.9523 |
| 1.0036 | 22.0 | 2249 | 0.7287 | 0.8698 |
| 0.7561 | 22.99 | 2351 | 0.5178 | 0.5407 |
| 0.6271 | 24.0 | 2454 | 0.3903 | 0.4185 |
| 0.5033 | 25.0 | 2556 | 0.2525 | 0.2356 |
| 0.4183 | 26.0 | 2658 | 0.1868 | 0.1624 |
| 0.351 | 26.99 | 2760 | 0.1435 | 0.1200 |
| 0.3256 | 28.0 | 2863 | 0.1154 | 0.1035 |
| 0.2859 | 29.0 | 2965 | 0.0996 | 0.0849 |
| 0.2771 | 29.93 | 3060 | 0.0996 | 0.0910 |
Framework versions
- Transformers 4.27.4
- Pytorch 2.0.0+cu118
- Datasets 2.11.0
- Tokenizers 0.13.3