Notes and exploration code for learning about AI/ML.
Multi-modal OCR pipeline optimized for ML training (text, figure, math, tables, diagrams).
Model Activity Visualiser for visualizing the internal workings of Large Language Models as they generate text.
Efficient full parameter tuning library for reinforcement learning applications in LLMs.
CodeScientist is an automated scientific discovery system for code-based experiments using LLMs.
基于OpenCompass的模型评测系统，提供用户自助开展评测工作的前端页面UI。
FlagEval is an evaluation toolkit for AI large foundation models.
Curated collection of system prompts for top AI tools, ideal for AI agent builders and prompt engineers.
A curated collection of MCP resources, guides, and servers for Chinese users.
Awesome-RAG-VIsion is a curated repository of advanced retrieval augmented generation techniques for Computer Vision.
A collection of high-quality pretrained models and resources for Chinese natural language processing.
Train a 26M-parameter visual multimodal VLM from scratch in just 1 hour, suitable for deep learning enthusiasts.
