SOLVED Running this v1.1 on llama.cpp
#3
by JeroenAdam - opened
Should I open an issue in the llama.cpp repo to get this working? Tried on latest llama.cpp and having this issue:
llama_model_load: loading tensors from '.\models\ggml-vicuna-13b-1.1-q4_0.bin'
llama_model_load: unknown tensor '�{ϻ��ꙛ|幷��dg�� ��?+��d�ȕw�eW8��' in model file
llama_init_from_file: failed to load model
Still in draft: https://github.com/ggerganov/llama.cpp/pull/896
It looks like you're not up to date. git pullthen make
Also try verifying sha256
JeroenAdam changed discussion title from Running this v1.1 on llama.ccp to Running this v1.1 on llama.cpp
JeroenAdam changed discussion title from Running this v1.1 on llama.cpp to SOLVED Running this v1.1 on llama.cpp