arthurwolf

• High-VRAM GPUS for us nerds.

  Posted by arthurwolf@reddit | LocalLLaMA | View on Reddit | 96 comments

• Cool tool for coding with LLMs: Prompt-Tower

  Posted by arthurwolf@reddit | LocalLLaMA | View on Reddit | 3 comments

• Llama.cpp PR adding support for Xeon Phi cards (240 cores), 4x inference speedup.

  Posted by arthurwolf@reddit | LocalLLaMA | View on Reddit | 11 comments