Jinja docker template

[michaelfeil]

This pull request includes significant updates to the Dockerfile setup and workflow configurations to enhance the build process for different platforms. The changes focus on adding support for NVIDIA, CPU, and AMD builds, improving the Dockerfile generation process, and refining the workflow for Docker container releases.

Dockerfile and Workflow Enhancements:

• Workflow Configuration Updates:

  • .github/workflows/release.yaml: Added separate jobs for NVIDIA and CPU Docker container pushes, including support for multiple platforms (linux/amd64, linux/arm64).
  • .github/workflows/release_docker_container.yaml: Introduced appendix_tag and platforms inputs to allow dynamic tagging and platform specification during Docker builds. [1] [2]

• Dockerfile Generation and Template Improvements:

  • libs/infinity_emb/Dockerfile.jinja2: Created a Jinja2 template for generating Dockerfiles for different platforms (NVIDIA, CPU, AMD), ensuring consistency and reducing manual edits.
  • libs/infinity_emb/docker.template.yaml: Added configuration for generating Dockerfiles using the Jinja2 template, specifying base images and installation commands for each platform.
  • libs/infinity_emb/Dockerfile, libs/infinity_emb/Dockerfile.amd_auto, libs/infinity_emb/Dockerfile.cpu_auto: Updated Dockerfiles to include autogenerated warnings and platform-specific configurations. [1] [2] [3]

• Makefile Enhancements:

  • libs/infinity_emb/Makefile: Added a template_docker target to automate the generation of Dockerfiles using the Jinja2 template and the configuration file. [1] [2]

These changes streamline the Docker build process, making it easier to manage and deploy containers across different platforms.

[greptile-apps]

PR Summary

This PR introduces a Jinja2-based templating system for generating Dockerfiles, improving flexibility and maintainability across different environments (CUDA, CPU, AMD).

• Added Dockerfile.jinja2 template and docker.template.yaml configuration file for customizable Docker builds
• Created Dockerfile.amd_auto and Dockerfile.cpu_auto for AMD and CPU-specific builds
• Updated pyproject.toml with new dependencies and Poetry sources for different PyTorch configurations
• Modified Makefile to include a new 'template_docker' target for generating Dockerfiles
• Changed import condition in infinity_emb/transformer/audio/torch.py to require both torch and transformers

_{9 file(s) reviewed, 24 comment(s)
Edit PR Review Bot Settings | Greptile}

[greptile-apps]

logic: PYTHON variable set to python3.11, but base image uses python3.10. This mismatch may cause issues.

[michaelfeil]

Fair, but amd is currently not working

[codecov-commenter]

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 79.15%. Comparing base (46ff710) to head (b866753).

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #426   +/-   ##
=======================================
  Coverage   79.15%   79.15%           
=======================================
  Files          41       41           
  Lines        3229     3229           
=======================================
  Hits         2556     2556           
  Misses        673      673           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.