MKRWW/Qwen3.8-27B-int4-AutoRound
43.6k
card: add reasoning/serving notes (thinking behavior, reasoning_content, max_tokens)
card: consistent weight size
Fix weight size in card (~18 GB)
Qwen3.8-27B int4 AutoRound (method-matched to 3.6, vLLM-ready)
initial commit
