Update README.md
Browse files
README.md
CHANGED
|
@@ -20,7 +20,14 @@ Its architecture is based on X-Codec with several major differences:
|
|
| 20 |
- **Semantic Supervision During Training**: It adds a semantic reconstruction loss, ensuring that the discrete tokens preserve meaningful linguistic and emotional information — crucial for TTS tasks.
|
| 21 |
- **Transformer-Friendly Design**: The 1D token structure of X-Codec2 naturally aligns with the autoregressive modeling in LLMs like LLaMA, improving training efficiency and downstream compatibility.
|
| 22 |
|
| 23 |
-
## Usage example
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 24 |
|
| 25 |
Here is a quick example of how to encode and decode an audio using this model:
|
| 26 |
|
|
|
|
| 20 |
- **Semantic Supervision During Training**: It adds a semantic reconstruction loss, ensuring that the discrete tokens preserve meaningful linguistic and emotional information — crucial for TTS tasks.
|
| 21 |
- **Transformer-Friendly Design**: The 1D token structure of X-Codec2 naturally aligns with the autoregressive modeling in LLMs like LLaMA, improving training efficiency and downstream compatibility.
|
| 22 |
|
| 23 |
+
## Usage example
|
| 24 |
+
|
| 25 |
+
Since Xcodec2 isn't yet merged into Transformers, you can install from source from the [corresponding fork](https://github.com/Deep-unlearning/transformers/tree/add-xcodec2).
|
| 26 |
+
|
| 27 |
+
Setup
|
| 28 |
+
```python
|
| 29 |
+
pip install git+https://github.com/Deep-unlearning/transformers.git@add-xcodec2
|
| 30 |
+
```
|
| 31 |
|
| 32 |
Here is a quick example of how to encode and decode an audio using this model:
|
| 33 |
|