Could you share the measurement.json file?
#1
by
Mukaisan
- opened
I have a GPU configuration of 12G+24G, so I’d like to create a quantized version that suits my setup.
Thank you so much for all your work!
I have a GPU configuration of 12G+24G, so I’d like to create a quantized version that suits my setup.
Thank you so much for all your work!
Uploaded the latest measurements here:
https://huggingface.co/LoneStriker/ExLlamaV2-Measurements