reforcement-learning

《高级机器学习》：面向生成式人工智能时代的机器学习教材，涵盖强化学习、大模型训练与对齐、生成建模和多模态学习等内容。

The implementation for our paper: TurnSight: Turn-Level Hindsight Self-Distillation for Tool-Integrated Reasoning.

Based on the virtual world built with IsaacSim, accomplish the training and deployment of real-world reinforcement learning for Hil-Serl.

Interactive Autonomous Navigation Research Lab for UGVs featuring localization, path planning, DWA, MPC, Adaptive MPC, reinforcement learning hooks, safety supervision, live visualization, benchmarking, replay, and automated report generation.

基于ms-swift的Qwen3_8B 金融推理两阶段后训练（LoRA SFT->GRPO）。

A learning path from Mechanical Engineering to AI and Robotics: 18 university course projects with source code and concept notes, for engineers moving into ML from a non-CS background. 从机械工程转 AI / 机器人学的开源学习路线 —— 本科到硕士的 18 个课程项目，含源码、原理讲解与工科知识映射表。