Accent Classification
Whisper-based audio classification model trained on the EdAcc dataset.
This model was trained for 15 epochs.
Evaluation
It achieves the following results on the evaluation set:
- Loss: 0.0340
- Accuracy: 0.9960
Hyperparameters
The following hyperparameters were used during training:
- learning_rate: 1e-05
- train_batch_size: 32
- eval_batch_size: 16
- seed: 0
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 15.0
- mixed_precision_training: Native AMP
Training Results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 2.75 | 1.0 | 290 | 2.6410 | 0.2788 |
| 1.7626 | 2.0 | 580 | 1.6589 | 0.6165 |
| 1.2434 | 3.0 | 870 | 1.0752 | 0.7746 |
| 0.903 | 4.0 | 1160 | 0.7628 | 0.8295 |
| 0.6878 | 5.0 | 1450 | 0.5528 | 0.8865 |
| 0.5063 | 6.0 | 1740 | 0.4051 | 0.9189 |
| 0.3312 | 7.0 | 2030 | 0.3041 | 0.9471 |
| 0.2562 | 8.0 | 2320 | 0.2236 | 0.9629 |
| 0.2163 | 9.0 | 2610 | 0.1633 | 0.9757 |
| 0.127 | 10.0 | 2900 | 0.1250 | 0.9813 |
| 0.1187 | 11.0 | 3190 | 0.0842 | 0.9888 |
| 0.0649 | 12.0 | 3480 | 0.0604 | 0.9920 |
| 0.0506 | 13.0 | 3770 | 0.0447 | 0.9945 |
| 0.0514 | 14.0 | 4060 | 0.0370 | 0.9956 |
| 0.0311 | 15.0 | 4350 | 0.0340 | 0.9960 |
Disclaimer
This model may make mistakes and should not be used in high-risk situations.
THE MODEL IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS MODEL INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS MODEL.
- Downloads last month
- 38