wav2vec2-large-xlsr-en-demo
This model is a fine-tuned version of jonatasgrosman/wav2vec2-large-xlsr-53-english on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.1356
- Wer: 0.2015
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: 1
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 5.3911 | 0.5 | 500 | 0.5397 | 0.2615 |
| 0.3413 | 1.01 | 1000 | 0.1423 | 0.2137 |
| 0.243 | 1.51 | 1500 | 0.1458 | 0.2210 |
| 0.2232 | 2.01 | 2000 | 0.1380 | 0.2143 |
| 0.162 | 2.51 | 2500 | 0.1464 | 0.2149 |
| 0.1384 | 3.02 | 3000 | 0.1348 | 0.2109 |
| 0.1164 | 3.52 | 3500 | 0.1324 | 0.2040 |
| 0.1103 | 4.02 | 4000 | 0.1310 | 0.2051 |
| 0.0857 | 4.53 | 4500 | 0.1356 | 0.2015 |
Framework versions
- Transformers 4.17.0
- Pytorch 1.12.1+cu113
- Datasets 1.18.3
- Tokenizers 0.12.1