3.5 - 4.0 bpw quant request

#1
by SporkySporkness - opened

Would love to have an exl2 to fit on 40gb vram with >20k context. I think 4bpw h6 or 3.5 h8. Thank you!

Hey @SporkySporkness , hope the day is well. Still need this? I can start the job.

Sign up or log in to comment