5bpw?

#4
by justj0sh - opened

Why was it skipped?

Patience. 1 more hour needed then upload time.

image

It was skipped because I didn't need it for the tuned quant so far.

But I need it for 128GiB (4x RTX 5090) and 144 GiB (6x 3090) folks, it's coming.

Thanks for your excitement and patience.

Available now https://huggingface.co/mratsim/GLM-4.7-EXL3/tree/5bpw_H8 !

mratsim changed discussion status to closed

Sign up or log in to comment