/UniAnimate-GradioUI

WebUI & Docker image of UniAnimate

Primary LanguagePython

image

News

  • 新增Docker_自动启动.bat批处理脚本，在你安装并正确配置了Git, CUDA, Docker的前提下，基本可以实现从Docker pull & git clone到浏览器WebUI页面一键启动的功能。

使用说明 [zh]

  1. 项目基于UniAnimate,我仅在此项目基础上做了Docker镜像的构建，以及借助Claude 3.5的共同协助，写了一个非完成体的WebUI脚本，您可以继续自行完善（例如：配置文件的修改保存方法）。

  2. 如果你安装的是Windows版本，也可以使用该UI脚本，但是请确保你的各目录路径设置正确。从头部署项目建议用pip install -r requirements2.txt,该依赖适配于python=3.10, CUDA=12.1, pytorch=2.2.2

  3. Docker的基础镜像基于pytorch/pytorch:2.2.2-cuda12.1-cudnn8-runtime这个官方镜像，你也可以自行选择其他版本的pytorch和CUDA。另外，镜像默认挂载卷D:\PythonProject\UniAnimate:/workspace/UniAnimate，你可以在启动容器时，自行修改挂载卷的路径。镜像使用方法：

    • a. 拉取镜像--> From:Docker hub
    docker pull aimaster104/unianimate:webui
    • 拉取镜像--> From:阿里云镜像
    另有阿里云私有镜像
    • b. 启动容器
    docker run --gpus all -p 7860:7860 --name UniAnimate-webui -v D:\PythonProject\UniAnimate-GradioUI:/workspace/UniAnimate -w /workspace/UniAnimate -d aimaster104/unianimate:webui python3 -m webui

    • c. 打开浏览器

      完成容器启动后，可以打开浏览器localhost:7860进入到浏览器WebUI中。

  4. WebUI的进度条显示仅供参考，具体执行进度和步骤还是以后台命令行显示的为准。（其实是因为不太擅长用Gradio, ^_^） image-1

image

  1. 最终视频生成结果在/outputs/UniAnimate_infer/文件夹中，你也可以自行修改生成结果的保存路径。
  2. 关于模型下载与放置目录请看官方说明. 另外，项目运行时，模型加载时间可能较长，请耐心等待。
  3. 如果你的显卡和内存硬件不够好（最好高于12G显存&16G内存），请选择用[512,768]的像素尺寸。

我的联系方式

我的抖音账号: AI大师工作流 BiliBili,小红书，今日头条，快手等短视频平台均同号。

Instructions for use [en]

  1. The project is based on UniAnimate. I only built the Docker image on this project and wrote a non-completed script of WebUI with the help of Claude 3.5. You can continue to improve it yourself (for example, the modification and saving method of the configuration file).

  2. You can also use this UI script if you are installing the Windows version, but make sure that your directory paths are set correctly. It is recommended to use pip install -r requirements2.txt for the project to deploy from scratch, which is adapted to python = 3.10, CUDA = 12.1, pytorch = 2.2.2

  3. This Docker image is based on pytorch/pytorch:2.2.2-cuda12.1-cudnn8-runtime，You can also choose other versions of pytorch and CUDA. In addition, the default mount volume of the mirror is D:\PythonProject\UniAnimate:/workspace/UniAnimate. You can modify the path of the mount volume when starting the container. Docker use method:

    • a. Docker pull
    docker pull aimaster104/unianimate:webui
    • b. Docker run
    docker run --gpus all -p 7860:7860 --name UniAnimate-webui -v D:\PythonProject\UniAnimate-GradioUI:/workspace/UniAnimate -w /workspace/UniAnimate -d aimaster104/unianimate:webui python3 -m webui
    • c. Open your browser

    When you have finished starting the container, you can open a browser and go to localhost:7860 WebUI.

  4. The progress bar of webui is displayed for reference only. The specific execution progress and steps are subject to those displayed on the back command line. (In fact, I am not good at using Gradio ^_^)

image-1

image

  1. The final video generation result is in/outputs/UniAnimate_infer/folder.
  2. AboutModels Downloading and saving,follow the official instructions.The models will take a long time to load, please wait patiently.
  3. If your graphics card and memory hardware are not good enough (Better than 12G VRAM & 16G RAM), please choose the pixel size of [512,768].

My social media

My social media account: AI Master Workflow
BiliBili, and other short video platforms are all the same account.

Disclaimer

This open-source model is intended for RESEARCH/NON-COMMERCIAL USE ONLY. We explicitly disclaim any responsibility for user-generated content. Users are solely liable for their actions while using the generative model. The project contributors have no legal affiliation with, nor accountability for, users' behaviors. It is imperative to use the generative model responsibly, adhering to both ethical and legal standards.