Overview:

I am trying to train some ML models on a huge dataset for my research. I have requested for a machine with a GPU from my university, they provided me with RTX 3070. I have installed rapidsai using following cmd on my WSL [Ubuntu 22 .04]. I am using Python 3.8.16.

conda create -n rapids-23.02 -c rapidsai -c conda-forge -c nvidia  \
    rapids=23.02 python=3.8 cudatoolkit=11.2 \
    jupyterlab tensorflow

Problem:

I am using pycaret library to train different models (classification models). In pycaret while setting data you can specify it to use GPU. So while setting use_gpu parameter to 'force' it is giving me following error:
ImportError: cuML is outdated or not found. Required version is >=22.10, got 3.0.0

Question:

How should I install cuML of required version?