Files
2025-10-20 19:28:39 +08:00

30 lines
1.3 KiB
Markdown
Raw Permalink Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

## Quickstart
### Start with Docker
```bash
docker build -t text2video-metax .
metax-docker run -it --gpus=[0] -v <downloaded_model>:<model_dir_in_container> text2video-metax
```
[metax-docker](https://developer.metax-tech.com/softnova/category?package_kind=Cloud&dimension=metax&chip_name=%E6%9B%A6%E4%BA%91C500%E7%B3%BB%E5%88%97&deliver_type=%E5%88%86%E5%B1%82%E5%8C%85&series_name=metax-docker) 是沐曦官方的 docker wrapper, 允许自动将 gpu 挂载到容器内. Base 镜像需自行联系沐曦官方获取.
### 支持模型
- https://modelscope.cn/models/iic/text-to-video-synthesis
- https://www.modelscope.cn/models/Wan-AI/Wan2.1-T2V-1.3B
### 测试程序
1. 准备输入数据集,可以参考示例`dataset.json`
2. 在docker镜像里运行测试程序会根据`dataset.json`内容,在`output`目录下生成视频文件。
```bash
python3 main.py --model "<model_dir_in_container>" --json "dataset.json" --results "results.json" --outdir "output" --device cuda --dtype fp16
```
## 测试结果
| | A100 平均生成时间(秒) | MetaX C500 平均生成时间(秒) |
|------|-------------------------|----------------------------|
| iic/text-to-video-synthesis | 12 | 16 |
| Wan-AI/Wan2.1-T2V-1.3B | 195.2339 | 303.4363 |