nick_asr_COMBO_v2
This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.4474
- Wer: 0.6535
- Cer: 0.2486
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: 5e-05
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
---|---|---|---|---|---|
0.3049 | 1.0 | 687 | 1.5013 | 0.7015 | 0.2607 |
0.2294 | 2.0 | 1374 | 1.5933 | 0.6693 | 0.2612 |
0.261 | 3.0 | 2061 | 1.6275 | 0.6985 | 0.2687 |
0.2658 | 4.0 | 2748 | 1.5568 | 0.6729 | 0.2581 |
0.1704 | 5.0 | 3435 | 1.5363 | 0.6650 | 0.2529 |
0.2537 | 6.0 | 4122 | 1.5764 | 0.6669 | 0.2542 |
0.2333 | 7.0 | 4809 | 1.5285 | 0.6596 | 0.2519 |
0.168 | 8.0 | 5496 | 1.4945 | 0.6571 | 0.2500 |
0.3263 | 9.0 | 6183 | 1.4968 | 0.6547 | 0.2510 |
0.3238 | 10.0 | 6870 | 1.4474 | 0.6535 | 0.2486 |
Framework versions
- Transformers 4.18.0
- Pytorch 1.11.0+cu113
- Datasets 2.1.0
- Tokenizers 0.12.1
- Downloads last month
- 3