Pinned Loading

1. Llamathon Llamathon Public

   🦙 Put your LLM server through a marathon — one command to benchmark any OpenAI-compatible endpoint (Ollama, vLLM, llama.cpp), with model auto-detection, real-workload sizing, and self-contained HTM…

   HTML 1

2. vaillant-rag vaillant-rag Public

   RAG over local documents with any OpenAI-compatible LLM — hybrid search, FAISS, re-ranking, streaming, FastAPI. Named after a Verdun war pigeon.

   Python 1

3. fingertune fingertune Public

   Pinch-controlled rhythm game: real-time webcam hand tracking (MediaPipe Hand Landmarker) judged to the millisecond, 100% in the browser.

   TypeScript 1