llm-reasoning

Democratizing Reinforcement Learning for LLMs

The RL Bridge for LLM-based Agent Applications. Made Simple & Flexible.

The foundation layer for agentic intelligence.

✨✨Latest Papers and Benchmarks in Reasoning with Foundation Models

Official code for NeurIPS2025 "Iterative Self-Incentivization Empowers Large Language Models as Agentic Searchers"

Utterly unelegant prompts for local LLMs, with scary results.

PipelineLLM 是一个系统性的大语言模型（LLM）后训练学习项目，涵盖从监督微调（SFT）到偏好优化（DPO）、强化学习（RLHF/PPO/GRPO）再到持续学习（Continual Learning)的完整技术栈。