Downloads · 30 days
16
5% of all-time downloads
asr-africa/wav2vec2-xls-r-akan-50-hours
wav2vec2-xls-r-akan-50-hours is a automatic speech recognition model from asr-africa. 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
16
5% of all-time downloads
All-time downloads
333
Public
Parameters
315M
19 GB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors1.3 GB · 100%
From the Hugging Face model README
This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on an unknown 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 |
|---|---|---|---|---|---|
| 9.033 | 3.4602 | 500 | 0.9218 | 0.8080 | 0.2812 |
| 1.1487 | 6.9204 | 1000 | 0.3705 | 0.3226 | 0.1013 |
| 0.7608 | 10.3806 | 1500 | 0.3547 | 0.3076 | 0.0964 |
| 0.6425 | 13.8408 | 2000 | 0.3585 | 0.2975 | 0.0938 |
| 0.5345 | 17.3010 | 2500 | 0.3868 | 0.3146 | 0.0995 |
| 0.4479 | 20.7612 | 3000 | 0.4395 | 0.3233 | 0.1004 |
| 0.3678 | 24.2215 | 3500 | 0.5173 | 0.3128 | 0.0985 |
| 0.3043 | 27.6817 | 4000 | 0.5529 | 0.3202 | 0.1009 |
| 0.2411 | 31.1419 | 4500 | 0.6129 | 0.3027 | 0.0948 |
| 0.2082 | 34.6021 | 5000 | 0.6554 | 0.3140 | 0.0984 |
| 0.1748 | 38.0623 | 5500 | 0.7193 | 0.3154 | 0.0977 |
| 0.1531 | 41.5225 | 6000 | 0.7602 | 0.3094 | 0.0958 |
| 0.1376 | 44.9827 | 6500 | 0.8043 | 0.3027 | 0.0946 |
| 0.1251 | 48.4429 | 7000 | 0.8027 | 0.3065 | 0.0949 |