2024–26IndependentGitHub
Open tools for running AI locally
GraphRAG-Local-UI made Microsoft’s GraphRAG usable with local models, with an indexing, prompt-tuning and query interface. It is one of several open tools I maintain for running retrieval and agents on your own hardware.
- Outcome
- 2,321 stars and 291 forks on GraphRAG-Local-UI, and a PyPI package other developers built on it
- Role
- Author and maintainer
- Stack
- Python · GraphRAG · Ollama · LM Studio · Gradio · FastAPI · MLX · Unsloth
- stars on GraphRAG-Local-UI
- 2,321
- stars on GraphRAG-Local-UI
- forks
- 291
- forks
- stars on Vodalus
- 195
- stars on Vodalus
- stars on MoA-Ollama-Chat
- 116
- stars on MoA-Ollama-Chat
Why local
Many teams cannot send their documents to a hosted model. When Microsoft released GraphRAG in mid-2024 it assumed OpenAI endpoints and a command line. I rebuilt the workflow around local models through Ollama and LM Studio and gave it an interface for indexing, prompt tuning, querying and visualizing the graph.
2,948
Other tools
Vodalus builds fine-tuning datasets grounded in RAG and Wikipedia and trains them with MLX and Unsloth. MoA-Ollama-Chat adapts the mixture-of-agents idea to local models, so several small models draft and one aggregates. creative-pipeline-automation is a proof of concept for a content supply chain: YAML campaign briefs in, brand and legal checks with automatic fixes, variants in three aspect ratios and several languages out.