elsung/blackwell-llm-toolkit
Empirical recipes, configs, and benchmarks for running modern open-weight LLMs on NVIDIA Blackwell GPUs (RTX PRO 6000, RTX 50-series). NVFP4 + TensorRT-LLM + vLLM + llama.cpp + LMCache, all verified on sm_120.
GitHub repository with 5 stars and 0 forks.
Language: Shell