Skip to content

Latest commit

 

History

History
78 lines (63 loc) · 1.44 KB

PREPARATION.md

File metadata and controls

78 lines (63 loc) · 1.44 KB
Raw

Preparation

Installation

Requirements
  • Linux (Windows is not officially supported)
  • Python 3.7+
  • PyTorch 1.8 or higher
  • CUDA 10.1 or higher
  • NCCL 2
  • GCC 4.9 or higher
Dependencies
  • dask
  • decord
  • future
  • fvcore
  • hickle
  • lpips
  • matplotlib
  • nni
  • netcdf4
  • numpy
  • opencv-python
  • packaging
  • pandas
  • scikit-image<=0.19.3
  • six
  • scikit-learn
  • timm>=0.5.4,<=0.6.11
  • torch
  • torchvision
  • tqdm
  • xarray==0.19.0

Clone the ST-SSPL-AVP repository

git clone https://github.com/zhenglab/ST-SSPL-AVP

Install the Python and PyTorch Environments

Install the corresponding versions of Python and PyTorch, and also setup the conda environment.

conda env create -f environment.yml
conda activate st_sspl_avp

Install the Dependency Packages

python setup.py develop

Dataset Preparation

  • Download the corresponding datasets of ERA5 via WeatherBench Github Repo.

  • Unzip and copy the dataset files to $ST-SSPL-AVP/data directory as following shows:

ST-SSPL-AVP
├── configs
└── data
    |── weather
    |   ├── 2m_temperature
    |   ├── 10m_u_component_of_wind
    |   ├── 10m_v_component_of_wind
    |   ├── relative_humidity
    |   ├── total_cloud_cover