<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
colab-demo
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: 3.9910
- Wer: 0.9714
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: 1000
- num_epochs: 30
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
0.1212 | 2.14 | 500 | 3.6706 | 1.0757 |
0.2303 | 4.27 | 1000 | 2.6849 | 1.0578 |
0.3003 | 6.41 | 1500 | 3.2261 | 1.0605 |
0.2705 | 8.55 | 2000 | 3.3483 | 1.0844 |
0.2178 | 10.68 | 2500 | 3.2000 | 1.0219 |
0.1875 | 12.82 | 3000 | 2.2454 | 1.0159 |
0.1792 | 14.96 | 3500 | 2.7510 | 0.9973 |
0.1477 | 17.09 | 4000 | 2.6716 | 0.9847 |
0.1232 | 19.23 | 4500 | 2.5939 | 0.9807 |
0.1051 | 21.37 | 5000 | 3.3308 | 0.9794 |
0.0847 | 23.5 | 5500 | 3.3430 | 0.9814 |
0.0809 | 25.64 | 6000 | 3.2566 | 0.9595 |
0.0642 | 27.78 | 6500 | 3.6392 | 0.9654 |
0.0566 | 29.91 | 7000 | 3.9910 | 0.9714 |
Framework versions
- Transformers 4.23.1
- Pytorch 1.12.1+cu113
- Datasets 1.18.3
- Tokenizers 0.13.1