wav2vec2-base-keyword-spotting
This model is a fine-tuned version of facebook/wav2vec2-base on the superb dataset. It achieves the following results on the evaluation set:
- Loss: 0.0746
- Accuracy: 0.9843
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: 32
- eval_batch_size: 32
- seed: 0
- gradient_accumulation_steps: 4
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 5.0
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.8279 | 1.0 | 399 | 0.6792 | 0.8558 |
| 0.2961 | 2.0 | 798 | 0.1383 | 0.9798 |
| 0.2069 | 3.0 | 1197 | 0.0972 | 0.9809 |
| 0.1757 | 4.0 | 1596 | 0.0843 | 0.9825 |
| 0.1607 | 5.0 | 1995 | 0.0746 | 0.9843 |
Framework versions
- Transformers 4.11.0.dev0
- Pytorch 1.9.1+cu111
- Datasets 1.12.1
- Tokenizers 0.10.3
- Downloads last month
- 16
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.