llama.cpp icon indicating copy to clipboard operation
llama.cpp copied to clipboard
Published 3 months ago • verified publisher icon ggml-org

quantize: improve pattern matching for allowed tensors

Open EAddario opened this issue 9 months ago • 1 comments

This PR implements @slaren's regex matching recommendation for allowed tensors. For example: --tensor-type attn=q4_k will now apply to all tensors named *attn*

EAddario avatar Apr 20 '25 09:04 EAddario

Apologies for shotgun approach @ggerganov / @slaren / @ngxson, I'm not sure what the proper process to request a review is. This PR addresses #12511 deficiencies. Happy to close or move to draft if it's not suitable for merging

EAddario avatar May 03 '25 06:05 EAddario