Request: Qwen3.5-122B-A10B-int4-AutoRound-abliterated
#7
by OpenSourceHustle - opened
If anyone can produce int4 version of this model it would be superb!
Qwen3.5-122B-A10B-int4-AutoRound-abliterated
Thank you so much!
I've already switched to a new architecture for the abliteration process. wangzhang/Qwen3.5-35B-A3B-abliterated was done with the new approach and works really well! The new wangzhang/Qwen3.5-122B-A10B-abliterated-v1 is coming soon β stay tuned π