Downloads · 30 days
6
11% of all-time downloads
nkkbr/whisper-large-v3-zatoichi-ja-EX-5
whisper-large-v3-zatoichi-ja-EX-5 is a automatic speech recognition model from nkkbr. 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
6
11% of all-time downloads
All-time downloads
53
Public
Parameters
1.5B
18.5 GB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors6.2 GB · 100%
From the Hugging Face model README
This model is a fine-tuned version of openai/whisper-large-v3 on the None 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 |
|---|---|---|---|---|
| 1.0521 | 0.1667 | 5 | 0.9421 | 105.1188 |
| 0.8992 | 0.3333 | 10 | 0.7033 | 124.1316 |
| 0.5669 | 0.5 | 15 | 0.5989 | 73.4918 |
| 0.5751 | 0.6667 | 20 | 0.5610 | 70.7495 |
| 0.6176 | 0.8333 | 25 | 0.5313 | 78.0622 |
| 0.4358 | 1.0 | 30 | 0.5352 | 88.7569 |
| 0.4308 | 1.1667 | 35 | 0.5154 | 72.5777 |
| 0.4131 | 1.3333 | 40 | 0.5000 | 64.3510 |
| 0.3883 | 1.5 | 45 | 0.4946 | 73.3090 |
| 0.2682 | 1.6667 | 50 | 0.4899 | 68.0987 |
| 0.3788 | 1.8333 | 55 | 0.4824 | 70.6581 |
| 0.3787 | 2.0 | 60 | 0.4763 | 64.4424 |
| 0.2895 | 2.1667 | 65 | 0.4846 | 65.5393 |
| 0.2466 | 2.3333 | 70 | 0.4927 | 65.0823 |
| 0.2541 | 2.5 | 75 | 0.4996 | 65.0823 |
| 0.2208 | 2.6667 | 80 | 0.5012 | 66.6362 |
| 0.2181 | 2.8333 | 85 | 0.4991 | 67.4589 |
| 0.1083 | 3.0 | 90 | 0.4935 | 67.8245 |
| 0.1607 | 3.1667 | 95 | 0.4990 | 67.9159 |
| 0.1528 | 3.3333 | 100 | 0.5090 | 66.9104 |
| 0.1396 | 3.5 | 105 | 0.5158 | 64.8080 |
| 0.1285 | 3.6667 | 110 | 0.5212 | 64.6252 |
| 0.1529 | 3.8333 | 115 | 0.5231 | 64.5338 |
| 0.0478 | 4.0 | 120 | 0.5228 | 64.4424 |