Skip to content

RELEASES.md

Latest commit

 

History

History
51 lines (38 loc) · 1.56 KB

RELEASES.md

File metadata and controls

51 lines (38 loc) · 1.56 KB
Raw

Release Notes

This project does not have a stable release process yet. Current builds should be treated as alpha developer previews.

Build A Local Bundle

Install llama.cpp on the build machine:

brew install llama.cpp

Package:

scripts/package-macos.sh --llama-server "$(which llama-server)"

Output:

dist/LLMRunner.app
dist/llmrunner

Before Publishing A Release

Checklist:

  • Run swift build -c release.
  • Run bash -n scripts/package-macos.sh.
  • Run python3 -m py_compile examples/python_chatbot.py.
  • Build dist/LLMRunner.app.
  • Start the packaged CLI with dist/llmrunner start.
  • Pull or reuse a small GGUF model with dist/llmrunner models pull tiny.
  • Verify dist/llmrunner models search tiny shows recommended files.
  • Verify curl http://127.0.0.1:8080/v1/models.
  • Verify API key auth with LLMRUNNER_API_KEY set.
  • Verify dist/llmrunner logs --lines 20 shows startup and request logs.
  • Verify one chat completion.
  • Verify one streaming chat completion with "stream": true.
  • Verify one text completion.
  • Verify one streaming text completion with "stream": true.
  • Verify one embeddings request.
  • Verify otool -L dist/LLMRunner.app/Contents/MacOS/llmrunner points libllama and libggml at @executable_path/../Resources/lib.
  • Verify codesign --verify --deep --strict dist/LLMRunner.app.
  • Stop the service with dist/llmrunner stop.

Notarization

The current package script uses ad-hoc signing. For public downloadable binaries, use a Developer ID certificate and Apple notarization before publishing.