xtuner

/

llava-phi-3-mini-gguf

like 133

Follow

xtuner 89

Hello!
Thanks for the training and the conversion, first of all.
I am getting gibberish on llama.cpp and M1.

optionarters are,iry, but,RESS, and thus,idor, but,iseconds, } and thus, },umm, }ività, }RESS, }RESS, } }RESS, } }abeth, androg, and thus,ershell,abeth, undert,RESS,abeth, }eign, and thus,eign, } }RESS, } }RESS,idor,clic,RESS, }bose, }ktion,iry,ución, }abeth,ktion,RESS,uminate, } }RESS, }RESS,RESS,ioned,RESS, }idor,abeth,uminate,RESS,uminate,RESS,RESS, } }annels, whereas,,otal,RESS,abeth,idor,ktion,ività, }abeth,RESS,ktion,abeth,RESS,idor, } and thusp,RESS, }RESS,idor,abeth,RESS,RESS,ershell,RESS,uminate, }abeth, }RESS,abeth,RESS,,'abeth,RESS, } },derr,idor, }RESS,iseconds,RESS,abeth, } }bose,eign, }derr, }issance,umm, } }derr, butRESS, } } } } }RESS, } }RESS,

./llava-cli -m ./models/llava-phi-3-mini/ggml-model-f16.gguf --mmproj ./models/llava-phi-3-mini/mmproj-model-f16.gguf -p 'Describe the image.' --image models/llava-v1.6-mistral-7b/S.jpg -c 4096

Have you tried on M1? Or maybe have any idea what the problem is?
I am on commit 928e0b70, latest as of right now.
Thank you very much!