cleversmoke

• Build Own Docker Image with llama.cpp and MTP

  Posted by cleversmoke@reddit | LocalLLaMA | View on Reddit | 9 comments

• Llama.cpp MTP with Qwen3.6 27B on Headless RTX 3090

  Posted by cleversmoke@reddit | LocalLLaMA | View on Reddit | 44 comments