Upload reade me
Browse files
README.md
CHANGED
|
@@ -5,20 +5,59 @@ language:
|
|
| 5 |
pipeline_tag: audio-to-audio
|
| 6 |
---
|
| 7 |
|
| 8 |
-
# HCodec-1.5 with adaptive frame rate
|
| 9 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
1. Install dependencies from requirement.txt via pypi
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
|
| 12 |
-
|
| 13 |
-
|
| 14 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 15 |
|
| 16 |
```bash
|
| 17 |
#!/bin/bash
|
|
|
|
| 18 |
python audio_tokenizer.py
|
| 19 |
```
|
| 20 |
|
| 21 |
-
## Optional configuration
|
| 22 |
+ Customize your testing options about adaptive frame rate
|
| 23 |
|
| 24 |
```yaml
|
|
|
|
| 5 |
pipeline_tag: audio-to-audio
|
| 6 |
---
|
| 7 |
|
| 8 |
+
# QuarkAudio-HCodec-1.5: A Unified Discrete Audio Tokenizer with adaptive frame rate for High-Fidelity, Multitask Audio Generation
|
| 9 |
+
|
| 10 |
+
<p align="center">
|
| 11 |
+
<a href="https://arxiv.org/pdf/2512.20151">
|
| 12 |
+
<img src="https://img.shields.io/badge/Paper-ArXiv-red.svg" alt="Paper">
|
| 13 |
+
</a>
|
| 14 |
+
<a href="https://github.com/alibaba/unified-audio/tree/main/QuarkAudio-UniSE">
|
| 15 |
+
<img src="https://img.shields.io/badge/GitHub-Code-green.svg" alt="GitHub">
|
| 16 |
+
</a>
|
| 17 |
+
<a href="https://github.com/alibaba/unified-audio/tree/main/QuarkAudio-HCodec/HCodec-1.5/">
|
| 18 |
+
<img src="https://img.shields.io/badge/Model-Hugging%20Face-yellow.svg" alt="Hugging Face">
|
| 19 |
+
</a>
|
| 20 |
+
<a href="https://www.modelscope.cn/models/QuarkAudio/QuarkAudio-HCodec/">
|
| 21 |
+
<img src="https://img.shields.io/badge/Model-%20%E9%AD%94%E6%90%AD-orange.svg" alt="ModelScope">
|
| 22 |
+
</a>
|
| 23 |
+
</p>
|
| 24 |
+
|
| 25 |
+
<p align="center">
|
| 26 |
+
<a href="https://arxiv.org/pdf/2512.20151"><img src="HCodec.jpg" width="70%" /></a>
|
| 27 |
+
</p>
|
| 28 |
+
|
| 29 |
+
|
| 30 |
+
## 🎯 Quick Start: Run Inference in 3 Minutes
|
| 31 |
+
## 1. Installation
|
| 32 |
1. Install dependencies from requirement.txt via pypi
|
| 33 |
+
2. Download pretrained weights from Huggingface 🤗: [QuarkAudio/HCodec-1.5-adaptive](https://huggingface.co/QuarkAudio/HCodec-1.5-adaptive) and save them to ./checkpoints/
|
| 34 |
+
3. confirm the `ckpt_path` in file `conf/config_adaptive_v3.yaml` is valid
|
| 35 |
+
|
| 36 |
+
|
| 37 |
+
### 2. Clone Repository
|
| 38 |
+
|
| 39 |
+
```bash
|
| 40 |
+
git clone https://github.com/alibaba/unified-audio.git
|
| 41 |
+
cd QuarkAudio-HCodec
|
| 42 |
+
```
|
| 43 |
|
| 44 |
+
### 3. Create a Conda environment and install dependencies
|
| 45 |
+
|
| 46 |
+
```bash
|
| 47 |
+
conda create -n unise python=3.10
|
| 48 |
+
conda activate unise
|
| 49 |
+
pip install -r requirements.txt
|
| 50 |
+
```
|
| 51 |
+
|
| 52 |
+
## 4. Tokenizer
|
| 53 |
|
| 54 |
```bash
|
| 55 |
#!/bin/bash
|
| 56 |
+
|
| 57 |
python audio_tokenizer.py
|
| 58 |
```
|
| 59 |
|
| 60 |
+
## 5. Optional configuration
|
| 61 |
+ Customize your testing options about adaptive frame rate
|
| 62 |
|
| 63 |
```yaml
|