llama runner process has terminated exit status 1

[dayskk]

用ollama run qwen2:7b 还是用curl模式，都返回错误llama runner process has terminated exit status 1

[dayskk]

复现步骤如下 conda create -n llm-cpp python=3.11 conda activate llm-cpp pip install --pre --upgrade ipex-llm[cpp] mkdir llama-cpp cd llama-cpp init-llama-cpp.bat init-ollama.bat set OLLAMA_NUM_GPU=999 set no_proxy=localhost,127.0.0.1 set ZES_ENABLE_SYSMAN=1 set SYCL_CACHE_PERSISTENT=1

ollama serve

用管理员权限运行的 然后另开一个也是管理员权限，ollama run qwen2:7b，就出现了 cpu如上面截图，这款cpu的i7 ,i5 这2个笔记本都不行

[dayskk]

[dayskk]

重新安装后错误又变了

[sgwhat]

hi @dayskk , could you please run the ENV-Check script in https://github.com/intel-analytics/ipex-llm/tree/main/python/llm/scripts and reply us with the results?

[biyuehuang]

After upgrade MTL iGPU driver to 5762 and install OneAPI 2024.2, it works.

## open miniforge prompt as administrator
conda activate llm2
cd C:\Program Files (x86)\Intel\oneAPI
setvars.bat
cd C:\Users\Admin\Documents\ollama2
init-ollama.bat

set OLLAMA_NUM_GPU=999
set no_proxy=localhost,127.0.0.1
set ZES_ENABLE_SYSMAN=1
ollama serve

## open another miniforge prompt as administrator
conda activate llm2
cd C:\Program Files (x86)\Intel\oneAPI
setvars.bat
cd C:\Users\Admin\Documents\ollama2

ollama pull qwen2:7b
curl http://localhost:11434/api/generate -d "{\"model\": \"qwen2:7b\",\"prompt\": \"Why is the sky blue?\",\"stream\": false}"

[dayskk]

安装oneapi 打开一个cmd窗口，用pip安装，这样安装的包约1.4G，比原来节省了20G+。 mkdir intel/dpct set PYTHONUSERBASE=c:/intel/dpct pip install --user mkl-dpcpp==2024.2 onednn==2024.2 dpcpp-cpp-rt==2024.2

然后到conda环境的窗口 conda activate your_env set PATH=c:\intel\dpct\Library\bin;%PATH%

此方法无效