Downloads · 30 days
9
17% of all-time downloads
KasuleTrevor/whisper-ln-afrivoice-20hr-v1r
whisper-ln-afrivoice-20hr-v1r is a automatic speech recognition model from KasuleTrevor. Use it when you need speech turned into text. It is set up for transformers. The card lists the license as apache-2.0.
should probably proofread and complete it, then remove this comment. --
Downloads · 30 days
9
17% of all-time downloads
All-time downloads
52
Public
Parameters
242M
15.5 GB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors967 MB · 100%
From the Hugging Face model README
This model is a fine-tuned version of openai/whisper-small on the AfriVoice dataset. It achieves the following results on the evaluation set:
More information needed
More information needed
More information needed
The following hyperparameters were used during training:
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|---|---|---|---|---|---|
| 1.8166 | 0.9959 | 122 | 1.1650 | 0.6027 | 0.2190 |
| 0.9195 | 2.0 | 245 | 0.8001 | 0.4432 | 0.1748 |
| 0.6827 | 2.9959 | 367 | 0.6824 | 0.3888 | 0.1499 |
| 0.5423 | 4.0 | 490 | 0.6289 | 0.3686 | 0.1521 |
| 0.4422 | 4.9959 | 612 | 0.6103 | 0.3643 | 0.1489 |
| 0.3461 | 6.0 | 735 | 0.6109 | 0.3572 | 0.1469 |
| 0.2603 | 6.9959 | 857 | 0.6298 | 0.3834 | 0.1635 |
| 0.1774 | 8.0 | 980 | 0.6615 | 0.3790 | 0.1638 |
| 0.1164 | 8.9959 | 1102 | 0.6981 | 0.3648 | 0.1493 |
| 0.0705 | 10.0 | 1225 | 0.7199 | 0.3720 | 0.1582 |
| 0.0443 | 10.9959 | 1347 | 0.7533 | 0.3722 | 0.1505 |
| 0.0279 | 12.0 | 1470 | 0.7849 | 0.3645 | 0.1491 |
| 0.019 | 12.9959 | 1592 | 0.8098 | 0.3602 | 0.1474 |
| 0.0134 | 14.0 | 1715 | 0.8265 | 0.3634 | 0.1445 |
| 0.0101 | 14.9959 | 1837 | 0.8512 | 0.3580 | 0.1443 |
| 0.0078 | 16.0 | 1960 | 0.8626 | 0.3636 | 0.1448 |