Enhance Docker Setup and Update Documentation for LivePortrait

[Sunwood-ai-labs]

Overview

This pull request introduces essential updates to the Docker environment and enhances the documentation in both English and Japanese for the LivePortrait project. These changes aim to simplify the setup process and improve accessibility for users.

Changes Included

1. Docker Compose Setup:

   • Added a new docker-compose.yml that configures the application with NVIDIA GPU support, ensuring the application can scale effectively and is easy to deploy.

2. Dockerfile Creation:

   • Created a Dockerfile that sets up a Python environment on a NVIDIA CUDA base image. This includes installing necessary system packages and Python libraries, preparing the system for high-performance operations with GPU support.

3. Python Requirements:

   • Defined all necessary Python libraries in requirements.txt, ensuring compatibility and functionality for the LivePortrait project. This includes libraries like PyTorch, OpenCV, and other multimedia processing tools.

4. Documentation Updates:

   • Japanese Documentation (docs/README_JP.md): Added comprehensive documentation in Japanese, detailing project features, setup instructions, and recent updates to cater to our Japanese-speaking users.
   • README Update (readme.md): Updated the main README file with Docker Compose instructions and added a link to the Japanese documentation to enhance accessibility and provide clear setup guidance.

Justification

These updates are crucial for maintaining an efficient development environment and ensuring that the project is accessible to a broader audience. The Docker enhancements will streamline project setup and deployment, particularly for environments requiring GPU resources. The documentation updates in Japanese will help in engaging a wider community by providing localized support and instructions.

Impact

The changes will not affect existing project functionalities but will significantly improve the setup process and user experience, particularly for new users who may be setting up the project for the first time.

[cleardusk]

@Sunwood-ai-labs Apologies for the delayed response. We've added a Chinese version of the README. It would be fantastic if you could help add Japanese support!
Regarding the Docker file, we’re not sure how to verify it yet, and we’ll need some time to do the check.

[DevilaN]

Unfortunately it is not building properly. I got:

=> ERROR [app 6/10] COPY requirements.txt .

EDIT: After fixing context issue and properly building I still got:

liveportrait-app-1  | 
liveportrait-app-1  | CUDA Version 12.0.1
liveportrait-app-1  | 
liveportrait-app-1  | Container image Copyright (c) 2016-2023, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
liveportrait-app-1  | 
liveportrait-app-1  | This container image and its contents are governed by the NVIDIA Deep Learning Container License.
liveportrait-app-1  | By pulling and using the container, you accept the terms and conditions of this license:
liveportrait-app-1  | https://developer.nvidia.com/ngc/nvidia-deep-learning-container-license
liveportrait-app-1  | 
liveportrait-app-1  | A copy of this license is made available in this container at /NGC-DL-CONTAINER-LICENSE for your convenience.
liveportrait-app-1  | 
liveportrait-app-1  | [22:15:37] Load appearance_feature_extractor done.   live_portrait_wrapper.py:35
liveportrait-app-1  |            Load motion_extractor done.               live_portrait_wrapper.py:38
liveportrait-app-1  |            Load warping_module done.                 live_portrait_wrapper.py:41
liveportrait-app-1  | [22:15:38] Load spade_generator done.                live_portrait_wrapper.py:44
liveportrait-app-1  |            Load stitching_retargeting_module done.   live_portrait_wrapper.py:48
liveportrait-app-1  | Could not load library libcudnn_cnn_infer.so.8. Error: libnvrtc.so: cannot open shared object file: No such file or directory
liveportrait-app-1 exited with code 139

So something is still not there yet...