Can't setup nVidia GPU under Win11

[nekiee13]

Using Anaconda, from virtual environment (Win11), I installed llamacpp by following procedure. No errors.
The nVidia driviers are detected by environment (nvcc --version), but when runinng locally, it runs on CPU - why?

(D:\LLM\LocalGPT\localGPT) D:\LLM\LocalGPT\localGPT>set CMAKE_ARGS=-DLLAMA_CUBLAS=on
(D:\LLM\LocalGPT\localGPT) D:\LLM\LocalGPT\localGPT>set FORCE_CMAKE=1
(D:\LLM\LocalGPT\localGPT) D:\LLM\LocalGPT\localGPT>pip install llama-cpp-python==0.1.83 --no-cache-dir
Successfully installed diskcache-5.6.3 llama-cpp-python-0.1.83

(D:\LLM\LocalGPT\localGPT) D:\LLM\LocalGPT\localGPT>nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2023 NVIDIA Corporation
Built on Tue_Aug_15_22:09:35_Pacific_Daylight_Time_2023
Cuda compilation tools, release 12.2, V12.2.140
Build cuda_12.2.r12.2/compiler.33191640_0

(D:\LLM\LocalGPT\localGPT) D:\LLM\LocalGPT\localGPT>python run_localGPT.py
2023-09-24 18:54:06,090 - INFO - run_localGPT.py:221 - Running on: cpu

The drivers are visible, so I did not set path to them prior to llamacpp installation.
REM set PATH=C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.2\bin;%PATH%

What did I do wrong and how to correct this? I got zero error on complete installation.

[nirodhie]

try putting the same commands under Powershell, I think it worked there but not on CMD

then remember to use:

python run_localGPT.py --device_type cuda

[nekiee13]

I try!

…

On Wed, 25 Oct 2023, 15:56 nirodhie, ***@***.***> wrote: try putting the same commands under Powershell, I think it worked there but not on CMD then remember to use: python run_localGPT.py --device_type cuda — Reply to this email directly, view it on GitHub <#523 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BB2KT6KEINZ2ER43MBZ2MLTYBELBDAVCNFSM6AAAAAA5FCKYR2VHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM3TGOBRHAYDG> . You are receiving this because you authored the thread.Message ID: ***@***.***>