wav2vec2-large-xlsr-grosman-53-aug-texts-exp-1
This model was trained from scratch on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3151
- Wer: 0.1967
- Cer: 0.1224
- Per: 0.1394
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: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 30
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer | Per |
|---|
| 0.2586 | 1.0 | 64 | 0.3151 | 0.1967 | 0.1224 | 0.1394 |
| 0.1103 | 2.0 | 128 | 0.3210 | 0.1966 | 0.1238 | 0.1413 |
| 0.1103 | 3.0 | 192 | 0.3371 | 0.2014 | 0.1247 | 0.1421 |
| 0.1065 | 4.0 | 256 | 0.3212 | 0.1975 | 0.1230 | 0.1400 |
| 0.0957 | 5.0 | 320 | 0.3197 | 0.1965 | 0.1224 | 0.1398 |
| 0.0957 | 6.0 | 384 | 0.3286 | 0.1968 | 0.1235 | 0.1406 |
| 0.0854 | 7.0 | 448 | 0.3400 | 0.1952 | 0.1227 | 0.1405 |
| 0.0863 | 8.0 | 512 | 0.3253 | 0.1936 | 0.1226 | 0.1403 |
| 0.0863 | 9.0 | 576 | 0.3439 | 0.1952 | 0.1227 | 0.1403 |
| 0.086 | 10.0 | 640 | 0.3514 | 0.1968 | 0.1236 | 0.1408 |
| 0.0779 | 11.0 | 704 | 0.3353 | 0.1982 | 0.1239 | 0.1412 |
Framework versions
- Transformers 4.28.0
- Pytorch 2.6.0+cu124
- Datasets 4.0.0
- Tokenizers 0.13.3
