Open Source RL Framework Miles v0.1 Released for LLM Training
ying11231 · x · 2026-08-19
Radixark team released Miles v0.1, an open-source reinforcement learning framework designed for LLMs and multimodal models. Miles aims to solve the challenge of debugging RL runs, ensuring correctness and hardware efficiency at scale. Developed over 9 months with 72 contributors and 1,326 commits, it has been battle-tested on frontier open models like Kimi K3, DeepSeek V4, and Qwen 3.8. The framework is currently powering production RL workloads at companies including Humansand, Modal, IBM, and Nebius AI.
Related event: RadixArk open-sources Miles v0.1, an RL training framework for LLMs(9 posts)→
More from coding & agent
- Dev uses Grok to clean up 3D printer firmware, catches it sneaking in nonsense code — yacineMTB · 2026-10-16
- Guardrails can make risky delegation work — but it takes a full-time team, says iandanforth — iandanforth · 2026-10-16
- What can you truly delegate? Only tasks with unambiguous, verifiable results — iandanforth · 2026-10-16
- Yacine: vibe-coded repos with 3,330 agent-written tests are all useless — yacineMTB · 2026-10-16
- Dev argues coding models now beat nearly all engineers in every branch of software — zack_overflow · 2026-10-16
- Q4-quantized coder model on 8-core CPU builds interactive solar system demo in one shot — IntrepidMindExplorer · 2026-10-16