Hugging Face's logo

Spaces:
lordofc
/
gpt
Runtime error

App Files Community
gpt / gpt4free /docs /build-workflow.md
LoRDxdd's picture
LoRDxdd
Add gpt4free API for Hugging Face
a4b70d9
preview code
|
raw
history blame contribute delete
3.52 kB

Build Workflow Documentation

This document explains the comprehensive build workflow for g4f that creates packages for multiple platforms and package managers.

Workflow Overview

The .github/workflows/build-packages.yml workflow automatically builds multiple package formats when a version tag is pushed to the repository.

Supported Package Formats

  1. PyPI Package - Python wheel and source distribution
  2. Windows Executable - Standalone .exe file built with Nuitka
  3. Linux Executable - Standalone binary for Linux systems built with Nuitka
  4. macOS Executable - Standalone binary for macOS systems built with Nuitka (x64 and ARM64)
  5. Debian Packages - .deb files for Ubuntu/Debian (amd64, arm64, armhf)
  6. WinGet Package - Windows Package Manager manifest
  7. Docker Images - Multi-architecture container images

Triggering a Build

To trigger a build, push a version tag to the repository:

git tag v1.2.3
git push origin v1.2.3

The workflow will:

  1. Detect the tag and extract the version
  2. Build all package formats in parallel
  3. Create a GitHub release with all artifacts
  4. Publish to PyPI (for releases)
  5. Generate WinGet manifest for Windows Package Manager

Manual Build Triggering

You can also manually trigger builds using the workflow_dispatch event:

  1. Go to the "Actions" tab in GitHub
  2. Select "Build All Packages" workflow
  3. Click "Run workflow"
  4. Optionally specify a version number

Package Locations

After a successful build, packages are available:

  • GitHub Releases: All executables and packages as release assets
    • Python packages (wheel and source distribution)
    • Standalone executables for Windows, Linux, and macOS
    • Debian packages for AMD64, ARM64, and ARMv7 architectures
    • WinGet manifest files
  • PyPI: pip install g4f
  • Docker Hub: docker pull hlohaus789/g4f:latest
  • WinGet: winget install g4f (after manifest approval)

Build Requirements

The workflow handles all dependencies automatically, but for local development:

  • Python 3.10+
  • Nuitka for executables (replaces PyInstaller)
  • Docker for container builds
  • dpkg-deb for Debian packages

Customizing Builds

Key files for customization:

  • g4f_cli.py - Entry point for executable builds
  • scripts/build-nuitka.sh - Nuitka build script for all platforms
  • scripts/build-deb.sh - Debian package build script
  • winget/manifests/ - WinGet package manifest templates
  • .github/workflows/build-packages.yml - Main workflow configuration

Version Handling

The workflow supports multiple version sources:

  1. Git tags (preferred for releases)
  2. Environment variable G4F_VERSION
  3. Manual input in workflow dispatch

Version must follow PEP 440 format for PyPI compatibility.

Troubleshooting

Common issues and solutions:

  1. Build fails: Check Python version compatibility and dependencies
  2. Version errors: Ensure version follows PEP 440 format
  3. Missing artifacts: Check if all build jobs completed successfully
  4. Docker push fails: Verify Docker Hub credentials are set in repository secrets

Security Notes

The workflow uses secure practices:

  • Trusted action versions
  • Environment isolation
  • Secret management for credentials
  • No hardcoded sensitive data

Contributing

To improve the build system:

  1. Test changes locally first
  2. Update documentation
  3. Consider backward compatibility
  4. Test with multiple Python versions