- Open Claude Code: Anthropic claude-code source code v2.1.88, supporting two LLM_PROVIDER modes: anthropic (Native) and openai_compat (OpenAI-Compatible).[github]
- YouPassNow: Real-Time Voice Multimodal Knowledge Base Q&A Assistant [demo]
- LizAgents: A collection of intelligent agents, including 🦄Vision Agent, 🐷Memory Agent, 🐮CSV Agent, 🐳PDF Agent, 🐯Search Agent and 🐸Chat Agent. Powered by OpenAI, LangChain, Streamlit, Selenium and GPT-4V-Act.[github] [demo]
- 🚅 LiteLLM - fix: handle missing 'index' in images response [github] [pr]
- Related Issue: [Bug]: OpenRouter + google/gemini-3-pro-image-preview: ValidationError - missing index field in images array [issue]
- 🦙 LlamaIndex - fix bugs and optimize the extraction prompt and parse_fn in the GraphRAG notebooks [github] [pr]
- 🦙 RAG-Anything - fix: replace
__del__with atexit to fix RAGAnything cleanup warning [github] [pr]
- BERT, ColBERT and ColPali
- EasyR1 + Verl + Ray + QwenVL + GRPO
- SFTTrainer Source Code Exploration
- QLoRA Code Implementation and Process Analysis
- Efficient GRPO Training with Unsloth
- Distributed Training
- Llama Source Code Exploration
- Transformer Source Code Exploration
- Multimodal Large Models
- Best Practices for Optimizing LLMs (Prompt Engineering, RAG and Fine-tuning)
- Lightweight Visualization Tool for Deep Learning: wandb
- Ollama User Guide
- Hugging Face and Transformers
- Fine-tuning
- LLM Leaderboard Platform
- Agent Protocol Stack: MCP, A2A, and AG-UI
- CrewAI: The Leading Multi-Agent Framework
- Mem0: A Memory Layer Framework for Empowering AI Agents
- Using n8n to Build a Telegram Personal Assistant
- AI Code Editor Source Code Exploration
- MCP Technical Overview
- LangChain Source Code Exploration
- LangChain: Building Powerful Applications with Large Language Models
- Agent & MultiAgent
- Prompt Engineering
- Claude 3.5 Sonnet: Computer use
- Building Conversational Applications with Streamlit