Update README.md (#1)
Browse files- Update README.md (3fdd45c77e8ba57559eaa8b39fddd20ae5503998)
Co-authored-by: 张少添 <tiansz@users.noreply.huggingface.co>
README.md
CHANGED
|
@@ -14,8 +14,20 @@ license: cc-by-4.0
|
|
| 14 |
|
| 15 |
This model was trained by kan-bayashi using csmsc/tts1 recipe in [espnet](https://github.com/espnet/espnet/).
|
| 16 |
### Demo: How to use in ESPnet2
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
```python
|
| 18 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19 |
```
|
| 20 |
### Citing ESPnet
|
| 21 |
```BibTex
|
|
|
|
| 14 |
|
| 15 |
This model was trained by kan-bayashi using csmsc/tts1 recipe in [espnet](https://github.com/espnet/espnet/).
|
| 16 |
### Demo: How to use in ESPnet2
|
| 17 |
+
You first need to import the following packages
|
| 18 |
+
```bash
|
| 19 |
+
pip install torch
|
| 20 |
+
pip install espnet_model_zoo
|
| 21 |
+
```
|
| 22 |
+
Then start using it!
|
| 23 |
```python
|
| 24 |
+
import soundfile
|
| 25 |
+
from espnet2.bin.tts_inference import Text2Speech
|
| 26 |
+
text2speech = Text2Speech.from_pretrained("espnet/kan-bayashi_csmsc_tts_train_tacotron2_raw_phn_pypinyin_g2p_phone_train.loss.best")
|
| 27 |
+
|
| 28 |
+
text = "春江潮水连海平，海上明月共潮生"
|
| 29 |
+
speech = text2speech(text)["wav"]
|
| 30 |
+
soundfile.write("out.wav", speech.numpy(), text2speech.fs, "PCM_16")
|
| 31 |
```
|
| 32 |
### Citing ESPnet
|
| 33 |
```BibTex
|