Skip to content

silasly/OpenMLR

 
 

Repository files navigation

OpenMLR Logo

OpenMLR

A self-hosted ML research agent that plans, researches, writes papers, and executes code — all in one conversation.

Deploy to Render

MIT License Quality Gate Status

Docs: openmlr.dev


Features

  • Plan + Execute modes — Plan mode gathers context; Execute mode does the work. Toggle with Cmd+M.
  • Paper research — OpenAlex, Semantic Scholar, arXiv, CrossRef, Papers With Code. Reads full papers, crawls citation graphs.
  • Paper writing — Section-by-section drafting with auto-save. Export to Markdown/LaTeX.
  • Compute environments — Execute code on local Docker, SSH remotes, or Modal cloud. Probe GPU/CPU capabilities.
  • Background jobs — Celery + Redis. Close the browser, come back later.
  • Multi-provider LLMs — OpenAI, Anthropic, OpenRouter, plus local models (Ollama, LM Studio).
  • MCP servers — Connect external tools via the Model Context Protocol.
  • Onboarding flow — Guided setup when no LLM provider is configured.

Quick Start

git clone https://github.com/xprilion/OpenMLR.git
cd OpenMLR
cp .env.example .env
make up
# or: docker compose -f docker-compose.prod.yml up -d

Open http://localhost:3000. Create an account. Add your API keys in Settings > Providers.

No API keys needed to start — the app guides you through configuration after login.

Development

Docker (recommended)

make dev-up         # Start with live reload
make dev-logs       # Watch logs

Native

make install        # Install dependencies
make infra          # Start Postgres + Redis in Docker
make db-fresh       # Create tables
make dev            # Start dev servers (backend :3000, frontend :5173)

Configuration

All LLM and tool API keys can be configured via the Settings UI after login. No environment variables are required to start.

For local development without Docker-managed databases:

DATABASE_URL="postgresql+asyncpg://user:pass@localhost:5432/openmlr"

See Configuration docs for all options.

Testing

make test           # All tests (backend + frontend + docs)
make test-backend   # Backend only
make test-frontend  # Frontend only
make lint           # Run linters

Architecture

frontend/   React 19 + TypeScript + Vite
backend/    Python 3.12 + FastAPI + SQLAlchemy + Celery
site/       VitePress documentation

See Architecture docs for details.

Contributing

  1. Fork the repo
  2. Create a feature branch
  3. Make your changes
  4. Run make test && make lint
  5. Submit a PR

License

MIT

About

A self-hosted ML research intern that plans tasks, reads papers, writes drafts, and runs experiments — end to end.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages

  • Python 66.2%
  • TypeScript 31.0%
  • Makefile 1.2%
  • CSS 0.8%
  • Dockerfile 0.2%
  • HTML 0.2%
  • Other 0.4%