This model was trained from scratch on the librispeech_asr dataset. It achieves the following results on the evaluation set:
- Loss: 6.9251
- Wer: 1.7858
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 3.0
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 6.6487 | 0.28 | 500 | 6.8354 | 1.4719 |
| 6.5662 | 0.56 | 1000 | 6.7877 | 0.9371 |
| 6.4309 | 0.84 | 1500 | 6.7640 | 1.1317 |
| 6.7123 | 1.12 | 2000 | 6.7907 | 1.9354 |
| 6.7547 | 1.4 | 2500 | 6.7830 | 1.8854 |
| 6.6726 | 1.68 | 3000 | 6.8211 | 1.9203 |
| 6.6538 | 1.96 | 3500 | 6.8444 | 1.8235 |
| 6.5693 | 2.24 | 4000 | 6.8873 | 1.8606 |
| 6.7234 | 2.52 | 4500 | 6.8649 | 1.8126 |
| 6.5104 | 2.8 | 5000 | 6.9251 | 1.7858 |
Framework versions
- Transformers 4.17.0.dev0
- Pytorch 1.10.2+cu113
- Datasets 1.18.3
- Tokenizers 0.11.0