Request for AWQ Quant of Qwen3.5-40B-Claude-4.6-Opus-Deckard-Heretic-Uncensored-Thinking-i1
#6
by 0xburakcelik - opened
Hi QuantTrio,
I hope you're doing well.
Would you be able to create AWQ quantizations (especially 4-bit and 5-bit) for this model?
Model:
mradermacher/Qwen3.5-40B-Claude-4.6-Opus-Deckard-Heretic-Uncensored-Thinking-i1-GGUF
It's a strong 40B reasoning + creative model (Claude Opus distilled + Deckard-Heretic tuning) and many people are interested in running it with vLLM. Currently only GGUF versions are available, and the GPTQ is quite heavy (~40GB) due to hybrid quantization.
AWQ versions would be greatly appreciated by the community.
Thank you in advance for your amazing work!
Best regards,
ToDo List +1