Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Docker(CUDA) Support For SHARK #2030

Open
wants to merge 3 commits into
base: main
Choose a base branch
from

Conversation

karthikkurella
Copy link

@karthikkurella karthikkurella commented Dec 8, 2023

Introduction of Dockerfile with CUDA Environment

  • CUDA Base Image: Utilizes nvidia/cuda:12.3.1-runtime-ubuntu22.04 as the base image to ensure compatibility with NVIDIA's CUDA.

  • Package Updates and Installations:

    • Runs apt update and apt upgrade for the latest packages.
  • Vulkan SDK Installation:

    • Adds the Vulkan SDK repository and installs the Vulkan SDK.
  • Python 3.11 Setup:

    • Installs Python 3.11 and sets it as the default Python version.
  • Virtual Environment for SHARK:

    • Sets up and ensures automatic activation of a virtual environment for SHARK, enhancing isolation and dependency management.
  • Interactive Shell Configuration:

    • Configures Docker to start with an interactive bash shell, allowing for manual intervention and debugging within the container.

Running Instructions

To build and run the Docker container based on this Dockerfile, use the following commands:

  1. Build the Docker Image:
    docker build -t your-image-name .
    
  2. Run the image:
    docker run --gpus all -it  --rm -p 8080:8080 your-image-name

@karthikkurella karthikkurella marked this pull request as draft December 8, 2023 19:59
@karthikkurella karthikkurella changed the title Docker Support + CUDA For SHARK Docker(CUDA) Support For SHARK Dec 8, 2023
@karthikkurella karthikkurella marked this pull request as ready for review December 8, 2023 21:40
@dan-garvey
Copy link
Member

dan-garvey commented Dec 12, 2023

Hey thanks for contributing. We're planning a pretty significant overhaul, so this may stop working in the near future, but I'm happy to commit it if anyone would find it useful.

I'm not very familiar with docker, so I'm unclear as to the reason to use a cuda docker as the base if you're installing vk?

@dan-garvey dan-garvey self-assigned this Dec 12, 2023
@karthikkurella
Copy link
Author

Thank you @dan-garvey, The reason is that docker:cuda image contains CUDA driver to interact with Nvidia GPU.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants