Pinned Loading

1. club-3090 club-3090 Public

   Community recipes for serving LLMs on RTX 3090. Multi-engine (vLLM, llama.cpp, SGLang) and model-agnostic. Currently shipping Qwen3.6-27B configs for 1× and 2× cards.

   Python 977 49

2. benchlocal-cli benchlocal-cli Public

   CLI port of BenchLocal quality bench packs — runs LLM behavioral evals against any OpenAI-compatible endpoint. Companion to club-3090.

   Python 2

3. vllm-project/vllm vllm-project/vllm Public

   A high-throughput and memory-efficient inference and serving engine for LLMs

   Python 80.3k 16.9k