Would you be open to redoing this model?
#2
by Shuasimodo - opened
If not, no prob and sorry to be a bother.
Just interested in how this model performs in both f16 and bf16 compared to existing quants based on the result of qwen3-coder-next.
I know most of the new qwen models are apparently better. But it's still a large a3b model and I'm curious to see how it would pair working alongside qwen3-coder-next.
Shuasimodo changed discussion status to closed