lmcoleman/Qwen3.8-27B-MagicQuant-GGUF

Upload Qwen3.8-27B-Q6_K.gguf with huggingface_hub

Upload Qwen3.8-27B-Q4_K_M.gguf with huggingface_hub

Card: state that throughput figures are CPU-only, not GPU

Card: document why no Q5 tier is published

Withdraw Q5 tier: dominated by Q4 on size, quality and speed

Upload mmproj-Qwen3.8-27B-f16.gguf

Upload Qwen3.8-27B-Q6_K.gguf

Upload Qwen3.8-27B-Q5_K_M.gguf

Upload Qwen3.8-27B-Q4_K_M.gguf

Upload README.md with huggingface_hub

initial commit