rllm-org / rllm
Democratizing Reinforcement Learning for LLMs
Why this job: says: reinforcement learning; topics: llm-training, reinforcement-learning
What it does
- Any harness. 10+ CLI harnesses (Claude Code, Codex, Terminus-2, mini-swe-agent, opencode, ...) plus Harbor-compatible task dirs. Or wrap your own agent — LangGraph, OpenAI Agents SDK, openai.OpenAI — with @rllm.rollout.
- Any sandbox. Docker, Daytona, Modal, or local. with snapshot + warm-pool acceleration to keep rollouts cheap at training-scale.
- Multiple training backends, one API. verl (distributed multi-GPU), tinker (single-machine), fireworks (Fireworks platform). Switch with one flag.
- 60+ integrated benchmarks. Math, code, MCQ, QA, search, VLM, translation, agentic — Terminal-Bench 2.0, SWE-bench, SkillsBench, AIME, MATH-500, GPQA, and more. rllm eval auto-pulls and runs.
- Multiple training methods. GRPO, REINFORCE, RLOO, SFT, on-policy distillation, and more.
Get started
uv pip install "rllm @ git+https://github.com/rllm-org/rllm.git" - Maintenance
- Active
- Last commit 32 days ago
- Stars
- 5,912
- +4 this week
- Licence
- Apache-2.0
- Allows business use
- Downloads for
- None listed
- No release files
- Latest tag
- v0.3.0-pre
- Released 6 months ago
- Built with
- Python
- 72% of the code
Is it looked after?
- Contributors
- 5
- Last 90 days
- Issues
- 71% closed
- 80 still open
- Pull requests
- 437 merged
- 102 waiting
- Releases
- 4
- 3 in the last year
- Contributing guide (opens in new tab)
- No security policy
- No code of conduct
Similar repos
| Repo | Stars | Last commit | Licence | Downloads for |
|---|---|---|---|---|
| aiming-lab / MetaClaw | 3,467 | 5 months ago | MIT | No downloads |
| Human-Agent-Society / reef | 7,794 | today | Apache-2.0 | No downloads |
More: languages, topics and project record
Languages
- Python 72%
- Jupyter Notebook 26%
- Shell 1%
- Go Template <1%
Topics
- agent-framework
- agentic-workflow
- coding-agent
- distributed-training
- llm-reasoning
- llm-training
- machine-learning
- ml-infrastructure
- ml-platform
- reinforcement-learning
- search-agent
- swe-agent
Project record
- JobFine-tuning
- StatusListed
- Commits1,926
- Forks638
- Watching30
- Stars in 28 days+40
- CreatedJan 26, 2025
- First seen by usNov 7, 2026
- CheckedNov 7, 2026