Upload FP16 checkpoint 920k: README.md
Browse files
README.md
CHANGED
|
@@ -20,7 +20,7 @@ Vietnamese zero-shot TTS / voice cloning fine-tuned from [ZipVoice](https://gith
|
|
| 20 |
- GitHub: https://github.com/iamdinhthuan/ViZipvoice
|
| 21 |
- Model repo: https://huggingface.co/contextboxai/ViZipvoice
|
| 22 |
- Space: https://huggingface.co/spaces/dinhthuan/ViZipvoice
|
| 23 |
-
- Latest checkpoint: `checkpoint-
|
| 24 |
- Training data: about `7000` total hours, including roughly `6500` hours of Vietnamese and `500` hours of English
|
| 25 |
- Tokenizer: `SimpleTokenizer`, character-level, `244` tokens
|
| 26 |
- Sample rate: `24 kHz`
|
|
@@ -129,7 +129,8 @@ Useful knobs:
|
|
| 129 |
|
| 130 |
## Files
|
| 131 |
|
| 132 |
-
- `checkpoint-
|
|
|
|
| 133 |
- `config.json`, `model.json`: model config
|
| 134 |
- `tokens.txt`: Vietnamese character tokenizer
|
| 135 |
- `audio/`: 30 reference audios plus `.txt` transcripts
|
|
|
|
| 20 |
- GitHub: https://github.com/iamdinhthuan/ViZipvoice
|
| 21 |
- Model repo: https://huggingface.co/contextboxai/ViZipvoice
|
| 22 |
- Space: https://huggingface.co/spaces/dinhthuan/ViZipvoice
|
| 23 |
+
- Latest checkpoint: `checkpoint-920000.pt`, FP16 inference state dict
|
| 24 |
- Training data: about `7000` total hours, including roughly `6500` hours of Vietnamese and `500` hours of English
|
| 25 |
- Tokenizer: `SimpleTokenizer`, character-level, `244` tokens
|
| 26 |
- Sample rate: `24 kHz`
|
|
|
|
| 129 |
|
| 130 |
## Files
|
| 131 |
|
| 132 |
+
- `checkpoint-920000.pt`: latest FP16 checkpoint
|
| 133 |
+
- `checkpoint-700000.pt`: previous FP16 checkpoint used for the current demo audios
|
| 134 |
- `config.json`, `model.json`: model config
|
| 135 |
- `tokens.txt`: Vietnamese character tokenizer
|
| 136 |
- `audio/`: 30 reference audios plus `.txt` transcripts
|