philbert440/Qwen3.6-40B-DeckardUncensored-OpusDistilled-W4A16-AWQ

Fix tokenizer: drop leftover truncation block, restore upstream tokenizer.json/tokenizer_config.json (fixes vision >1024 tokens; see discussion #3)

Standardize model card

Add packed W4A16 MTP head (config.json)

Add packed W4A16 MTP head (model.safetensors.index.json)

Add packed W4A16 MTP head (model-mtp.safetensors)

Reference 1Cat-vLLM PR #88 V100/SM70 recommended serving config

Switch to MSE observer to eliminate deterministic typo class (canary 0/12 -> typo-free)

Update README.md

Update README.md

Update README.md

Create README.md

Add files using upload-large-folder tool

initial commit