metadata
tags:
- generated_from_trainer
datasets:
- librispeech_asr
model-index:
- name: ''
results: []
This model was trained from scratch on the librispeech_asr dataset. It achieves the following results on the evaluation set:
- Loss: 1.8365
- Wer: 0.2812
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: 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: 1000
- num_epochs: 20.0
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 5.8017 | 1.68 | 1500 | 5.7161 | 1.3220 |
| 4.5907 | 3.36 | 3000 | 4.7936 | 0.9799 |
| 3.151 | 5.04 | 4500 | 4.1610 | 0.7752 |
| 1.5166 | 6.73 | 6000 | 3.5939 | 0.5343 |
| 2.4523 | 8.41 | 7500 | 4.0013 | 0.6954 |
| 1.423 | 10.09 | 9000 | 2.6917 | 0.4476 |
| 0.7882 | 11.77 | 10500 | 2.4493 | 0.3967 |
| 1.1643 | 13.45 | 12000 | 2.0629 | 0.3234 |
| 0.5352 | 15.13 | 13500 | 2.0625 | 0.3363 |
| 0.407 | 16.82 | 15000 | 1.8378 | 0.2812 |
| 0.1162 | 18.5 | 16500 | 1.8365 | 0.2812 |
Framework versions
- Transformers 4.17.0.dev0
- Pytorch 1.10.2+cu113
- Datasets 1.18.3
- Tokenizers 0.11.0