Phi3 3.8B mini 128k model not supported

[lumurillo]

I got an error when trying to load the Phi3 3.8B mini 128k model using the ipex-llm's support for llama.cpp:

The model that I tried is this: https://ollama.com/library/phi3:3.8b-mini-128k-instruct-q4_0

[sgwhat]

Hi @lumurillo , we have reproduced this issue and will inform you when we make progress.

[qiuxin2012]

@lumurillo We have fixed this issue, please update your ipex-llm to latest nightly build, and try again.