custom
File size: 4,457 Bytes
26f2a99
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
89f695e
26f2a99
 
 
 
 
 
 
 
89f695e
26f2a99
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
89f695e
26f2a99
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9e54747
26f2a99
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
---
license: mit
---
# X-VC

[![arXiv](https://img.shields.io/badge/arXiv-2604.12456-b31b1b.svg)](https://arxiv.org/abs/2604.12456)
[![GitHub](https://img.shields.io/badge/GitHub-Code-181717?logo=github)](https://github.com/Jerrister/X-VC)
[![Demo Page](https://img.shields.io/badge/Demo-Project%20Page-blue)](https://x-vc.github.io)

Official code release for **X-VC: Zero-shot Streaming Voice Conversion in Codec Space**.

## Environment Setup

### 1. Clone

```bash
git clone https://github.com/Jerrister/X-VC.git
cd X-VC
```

### 2. Create conda environment and install dependencies

```bash
conda create -n xvc python=3.10 -y
conda activate xvc
pip install -U pip
pip install -r requirements.txt
```

### 3. Prepare pretrained models

Prepare:
- [GLM-4-Voice-Tokenizer](https://huggingface.co/zai-org/glm-4-voice-tokenizer) (for semantic tokenization)
- [ERes2Net speaker encoder](https://modelscope.cn/models/iic/speech_eres2net_sv_en_voxceleb_16k) (for speaker feature extraction)

Then set paths in [`configs/xvc.yaml`](configs/xvc.yaml), especially:
- `model.generator.semantic_encoder.encoder.from_pretrained`
- `model.generator.semantic_encoder.cfg`
- `model.generator.speaker_encoder.pretrained_dir`

### 4. Prepare checkpoints

Put checkpoints under `ckpts/`, for example:

```text
ckpts/
  xvc.pt
```

## Inference

### Single-pair Inference

Use [`scripts/infer_single.sh`](scripts/infer_single.sh).

```bash
bash scripts/infer_single.sh
```

Key arguments in this script:
- `current=0` for offline inference.
- `current>0` for streaming inference.
- `chunk/current/future/smooth` control streaming behavior.

Outputs are saved under `save_dir` (default: `outputs/xvc_single`).

### Batch Offline Inference (SeedTTS-eval as example)

Use [`scripts/batch_infer_seedtts_offline.sh`](scripts/batch_infer_seedtts_offline.sh).

```bash
bash scripts/batch_infer_seedtts_offline.sh
```

This script reports:
- `saved_dir`
- `total_rtf`

### Batch Streaming Inference (SeedTTS-eval as example)

Use [`scripts/batch_infer_seedtts_stream.sh`](scripts/batch_infer_seedtts_stream.sh).

```bash
bash scripts/batch_infer_seedtts_stream.sh
```

This script reports:
- `saved_dir`
- `avg_latency_ms`

## Training

### Step 1: Prepare pretrained dependencies

Before training, prepare the required pretrained dependencies:
- [SAC pretrained checkpoint(s)](https://huggingface.co/Soul-AILab/SAC-16k-62_5Hz) (for model initialization)

Then set corresponding paths in [`configs/xvc.yaml`](configs/xvc.yaml), especially:
- `model.generator.checkpoint`
- `model.discriminator.checkpoint`

### Step 2: Prepare training data

Organize your training/validation data in JSONL format and set:
- `datasets.train`
- `datasets.val`

in [`configs/xvc.yaml`](configs/xvc.yaml).

### Step 3: Modify training configs

You can adjust training behavior in:
- [`configs/xvc.yaml`](configs/xvc.yaml) (main training config)
- [`configs/ds_stage2.json`](configs/ds_stage2.json) (DeepSpeed config)

### Step 4: Start training

Use [`scripts/train.sh`](scripts/train.sh).

```bash
bash scripts/train.sh
```

Notes:
- Default training engine is DeepSpeed (`configs/ds_stage2.json`).
- Main experiment config is `configs/xvc.yaml`.
- Set your `WANDB_API_KEY` in `scripts/train.sh` before running if you use wandb logging.

## Data Format

Training config points to JSONL files in `configs/xvc.yaml`:
- `datasets.train`
- `datasets.val`

Each JSONL line should be a JSON object.

Required fields:
- `target_utt`
- `source_wav_path`
- `target_wav_path`

Optional field:
- `source_utt`

Minimal example:

```json
{"source_utt":"utt_0001","source_wav_path":"<path_to_source>","target_utt":"utt_0002","target_wav_path":"<path_to_target>"}
```

## Acknowledgements

This codebase builds upon open-source components from [SAC](https://github.com/Soul-AILab/SAC) and the broader audio generation ecosystem.

## Citation

If you find our work useful in your research, please consider citing:

```bibtex
@misc{zheng2026xvczeroshotstreamingvoice,
      title={X-VC: Zero-shot Streaming Voice Conversion in Codec Space}, 
      author={Qixi Zheng and Yuxiang Zhao and Tianrui Wang and Wenxi Chen and Kele Xu and Yikang Li and Qinyuan Chen and Xipeng Qiu and Kai Yu and Xie Chen},
      year={2026},
      eprint={2604.12456},
      archivePrefix={arXiv},
      primaryClass={eess.AS},
      url={https://arxiv.org/abs/2604.12456},
}
```
## License

This project is licensed under the MIT License.