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

Add Podman + Wasm + GPU section #223

Merged
merged 1 commit into from
Apr 2, 2024

Conversation

CaptainVincent
Copy link
Member

@CaptainVincent CaptainVincent commented Apr 2, 2024

Explanation

Add podman execution version (full steps).

Related issue

None, internal request.

What type of PR is this

Documentation.

Proposed Changes

Add other container manager clients support is more user-friendly, especially since support for CDI toolchain is not yet fully mature. This also records a tested and runnable combination for end user.

Copy link
Collaborator

alabulei1 commented Apr 2, 2024

Hello, I am a code review bot on flows.network. Here are my reviews of code commits in this PR.


Commit 80c829c3b0f0843a93436e0cbd06d82937f80514

Key Changes:

  1. Addition of a new section titled "Podman + WASM + GPU" in the documentation.
  2. Provides instructions for setting up GPU support using Podman, Crun with Wasmedge, and CDI for host GPU device usage.
  3. Includes prerequisite information such as having a GPU device (specifically tested with NVIDIA GPUs on Linux), installing the NVIDIA GPU driver, NVIDIA Container Toolkit, or nvidia-container-toolkit-base package, and needing Podman version 4.x or higher.
  4. Offers guidance on installation and verification of NVIDIA driver, toolkit, and Podman.
  5. Provides steps for generating the CDI specification file for host GPU device mapping.
  6. Instructions for setting up the container runtime (crun with wasmedge) and enabling the plugin system.
  7. Includes commands for downloading the ggml plugin and demo application for running inference models with GPU acceleration.

Potential Problems:

  1. No mention of specific operating system compatibility where these steps can be applied, assuming it's for Linux due to the NVIDIA GPU drivers and toolkit installation instructions.
  2. Reliance on Linuxbrew for Podman installation might limit the audience as Linuxbrew is not a widely adopted package manager.
  3. Lack of specific details on setting up crun with wasmedge may lead to confusion for users unfamiliar with these tools.
  4. The complexity of the setup involving manual compilation and installation of components may deter users looking for a more straightforward GPU + WASM setup.
  5. Incomplete information for running the demo, such as missing information on how to obtain the podman and crun paths.

Copy link
Member

@hydai hydai left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We should have a complete guide, not a simplified one.

@CaptainVincent
Copy link
Member Author

We should have a complete guide, not a simplified one.

Ok, I will add those part.

@CaptainVincent CaptainVincent marked this pull request as draft April 2, 2024 06:46
Most parts/steps are duplicate with the Docker version, so this is a simplified version of Podman,
providing only the necessary instructions and minimum version requirements.

Signed-off-by: vincent <[email protected]>
@CaptainVincent CaptainVincent marked this pull request as ready for review April 2, 2024 09:51
@CaptainVincent CaptainVincent requested a review from hydai April 2, 2024 09:52
@hydai hydai merged commit 76e788c into WasmEdge:main Apr 2, 2024
6 checks passed
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.

3 participants