wav2vec2-large-lv60-self-colab
This model is a fine-tuned version of facebook/wav2vec2-large-960h-lv60-self on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0563
- Wer: 0.1541
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: 8
- 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: 12
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
6.3084 | 1.0 | 500 | 3.0693 | 1.0 |
1.5665 | 2.01 | 1000 | 0.0834 | 0.1749 |
0.2761 | 3.01 | 1500 | 0.0620 | 0.1593 |
0.178 | 4.02 | 2000 | 0.0667 | 0.1577 |
0.1225 | 5.02 | 2500 | 0.0616 | 0.1587 |
0.1423 | 6.02 | 3000 | 0.0581 | 0.1542 |
0.1019 | 7.03 | 3500 | 0.0616 | 0.1559 |
0.0914 | 8.03 | 4000 | 0.0690 | 0.1553 |
0.0781 | 9.04 | 4500 | 0.0599 | 0.1541 |
0.0791 | 10.04 | 5000 | 0.0607 | 0.1544 |
0.0772 | 11.04 | 5500 | 0.0563 | 0.1541 |
Framework versions
- Transformers 4.24.0
- Pytorch 1.13.0+cu116
- Datasets 1.18.3
- Tokenizers 0.13.2
- Downloads last month
- 1
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.