wav2vec2-base-librispeech100h-google-colab
This model is a fine-tuned version of facebook/wav2vec2-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.1156
- Wer: 0.0756
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: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 1.6033 | 0.9 | 1600 | 0.4802 | 0.2728 |
| 0.1912 | 1.79 | 3200 | 0.1601 | 0.1140 |
| 0.1409 | 2.69 | 4800 | 0.1423 | 0.0932 |
| 0.108 | 3.59 | 6400 | 0.1260 | 0.0806 |
| 0.1045 | 4.48 | 8000 | 0.1156 | 0.0756 |
Framework versions
- Transformers 4.20.0.dev0
- Pytorch 1.11.0+cu113
- Datasets 2.2.3.dev0
- Tokenizers 0.12.1