This repo will showcase how we can use some modern PPLs to model extreme values for climate data. We will also investigate how we can extend the standard models for EVTs which include spatial and/or temporal considerations.
Clone Repo
git clone https://github.com/jejjohnson/jbayesevt.git
cd jbayesevt
Create Environment
conda env create -f environments/environment_ai_gpu.yaml -y
conda activate bayesevt_ai
Install Additional Packages There are a few libraries that do not install correctly via the conda environment.
pip install git+https://github.com/NVIDIA/apex.git -v --disable-pip-version-check --no-cache-dir --no-build-isolation --global-option="--cpp_ext" --global-option="--cuda_ext"