whisper-small-keyword-spotting-m
This model is a fine-tuned version of openai/whisper-small on the kw-spotting-fsc-sl-agv dataset. It achieves the following results on the evaluation set:
- Loss: 0.0015
- Accuracy: 0.9999
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: 0.001
- 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 |
|---|---|---|---|---|
| 0.0146 | 1.0 | 356 | 0.0231 | 0.9925 |
| 0.0124 | 2.0 | 712 | 0.0105 | 0.9977 |
| 0.0091 | 3.0 | 1068 | 0.0015 | 0.9999 |
| 0.0101 | 4.0 | 1425 | 0.0028 | 0.9994 |
| 0.0094 | 5.0 | 1780 | 0.0022 | 0.9995 |
Framework versions
- Transformers 4.29.0.dev0
- Pytorch 2.0.0
- Datasets 2.10.1
- Tokenizers 0.13.2