<!-- 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-Arabic-phoneme-Experiment3
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: 0.0054
- Per: 0.0179
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: 6
- 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: 30.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Per |
|---|---|---|---|---|
| 11.0415 | 1.0 | 102 | 3.1505 | 1.0 |
| 3.1122 | 2.0 | 204 | 3.0968 | 1.0 |
| 3.0896 | 3.0 | 306 | 3.0042 | 1.0 |
| 2.953 | 4.0 | 408 | 2.8056 | 1.0 |
| 2.444 | 5.0 | 510 | 1.7510 | 0.6859 |
| 1.1295 | 6.0 | 612 | 0.3595 | 0.1672 |
| 0.3895 | 7.0 | 714 | 0.1119 | 0.0782 |
| 0.2175 | 8.0 | 816 | 0.0607 | 0.0477 |
| 0.1353 | 9.0 | 918 | 0.0317 | 0.0247 |
| 0.0946 | 10.0 | 1020 | 0.0288 | 0.0286 |
| 0.0835 | 11.0 | 1122 | 0.0291 | 0.0370 |
| 0.0673 | 12.0 | 1224 | 0.0231 | 0.0261 |
| 0.0583 | 13.0 | 1326 | 0.0181 | 0.0199 |
| 0.0444 | 14.0 | 1428 | 0.0190 | 0.0266 |
| 0.041 | 15.0 | 1530 | 0.0134 | 0.0254 |
| 0.0357 | 16.0 | 1632 | 0.0122 | 0.0233 |
| 0.0301 | 17.0 | 1734 | 0.0098 | 0.0338 |
| 0.0265 | 18.0 | 1836 | 0.0099 | 0.0241 |
| 0.0241 | 19.0 | 1938 | 0.0098 | 0.0207 |
| 0.0258 | 20.0 | 2040 | 0.0089 | 0.0203 |
| 0.0234 | 21.0 | 2142 | 0.0104 | 0.0242 |
| 0.0193 | 22.0 | 2244 | 0.0126 | 0.0301 |
| 0.0186 | 23.0 | 2346 | 0.0073 | 0.0274 |
| 0.0155 | 24.0 | 2448 | 0.0073 | 0.0238 |
| 0.0143 | 25.0 | 2550 | 0.0056 | 0.0164 |
| 0.0134 | 26.0 | 2652 | 0.0060 | 0.0199 |
| 0.012 | 27.0 | 2754 | 0.0052 | 0.0206 |
| 0.0122 | 28.0 | 2856 | 0.0054 | 0.0179 |
| 0.0105 | 29.0 | 2958 | 0.0055 | 0.0188 |
| 0.0098 | 30.0 | 3060 | 0.0054 | 0.0179 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.0.1+cu118
- Datasets 1.18.3
- Tokenizers 0.13.3
