Tried loading A Phixtral model and got this error

LostRuins / koboldcpp

A simple one-file way to run various GGML and GGUF models with KoboldAI's UI

https://github.com/lostruins/koboldcpp

GNU Affero General Public License v3.0

4.35k stars 312 forks source link

Tried loading A Phixtral model and got this error #849

Open DjagbleyEmmanuel opened 1 month ago

DjagbleyEmmanuel commented 1 month ago

llama_model_load: error loading model: check_tensor_dims: tensor 'blk.0.ffn_down.weight' not found llama_load_model_from_file: failed to load model fish: Job 1, './koboldcpp-linux-x64-nocuda_new' terminated by signal SIGSEGV (Address boundary error)

LostRuins commented 1 month ago

Where did you get that model? What model is it?

DjagbleyEmmanuel commented 1 month ago

From hugging face Of course "https://huggingface.co/ugriffo/phixtral-2x2_8-GGUF"