This guide provides step-by-step instructions to set up the DSE-CBM (a speaker diarization toolkit driven by Pyannote(https://github.com/pyannote/pyannote-audio)) project in a Conda environment on a Linux system with CUDA support.
Update Conda to ensure you have the latest version:
conda update -n base -c defaults condaCreate a new Conda environment named dsecbm with Python 3.12:
conda create -n dsecbm python=3.12
conda activate dsecbmConfigure environment variables for CUDA 11.8 by adding the following lines to ~/.bashrc:
export CUDA_HOME="/usr/local/cuda-11.8/"
export PATH=$PATH:$CUDA_HOME/bin
export LD_LIBRARY_PATH=/usr/local/cuda-11.8/lib64:$LD_LIBRARY_PATHApply the changes:
source ~/.bashrcInstall CUDA Toolkit 11.8 and PyTorch with its dependencies:
conda install cudatoolkit=11.8 -c nvidia
pip install torch==2.4.1 torchvision==0.19.1 torchaudio==2.4.1 --index-url https://download.pytorch.org/whl/cu118git clone https://github.com/state-spaces/mamba.git
cd mamba/pip install .git clone https://github.com/lz-hust/DSE-CBM.git
cd DSE-CBM/
pip install -r requirements.txt- The code in this repository is licensed under the MIT license.