harness-hub releases first RL-trained harness adapter for agentic coding, full recipe coming Friday
yb2698 · x · 2026-11-17
The harness-hub team on Hugging Face released its first harness adapter (e.g. Qwen3.8-27B-pi-adapted), a LoRA trained with safe RL to make models more efficient inside specific coding-agent harnesses.
Key points:
- The LoRA runs on transformers, vLLM, sglang, llama.cpp, and MLX on macOS, or can be merged into the base model
- More harness adapters are coming in the next few days; a comprehensive blogpost with the full recipe drops Friday so anyone can reproduce and train the next generation of adapters
- The team previously shipped TwIL-8B adapters and the traj2envs RL environment
Related event: Harness Adapters Launch: LoRA Boosts Qwen3.8-27B Coding Agent Performance(3 posts)→
More from coding & agent
- Mnemos x NousResearch launches agent memory engine claiming 2x gains over Claude Code — RileyRalmuto · 2026-11-17
- How Revid doubled its ChatGPT-sent signups from 355 to 781 per week in 10 weeks — tibo_maker · 2026-11-17
- FormMCP: one MCP connection to a family of desktop apps, tools loaded on demand — nodemode_ai · 2026-11-17
- Harness Adapters: RL-trained LoRA lifts Qwen3.8-27B on Pi from 29% to 63% on SWE-bench Pro — LysandreJik · 2026-11-17
- Pi harness adapter ships as a LoRA runnable on vLLM, llama.cpp, and MLX — yb2698 · 2026-11-17
- Real SWE-Bench Pro task: base model hits 100-step limit, adapter solves it in 8 calls — yb2698 · 2026-11-17