conda install --yes --file requirements.txt
2.1 Get the data. This is a preprocessed dataset generously provided by Prof. Dimitrios Giannakis. It is not contained due to size issue, and requests for CCSM4 dataset should be addressed to lingdaw2@illinois.edu
2.2 Train the ConvGRU network under the dir ConvGRU_CCSM4
, and the trained models will be saved in ConvGRU_CCSM4/results/saved_ConvGRU_model
CUDA_VISIBLE_DEVICES=0 python main.py
CUDA_VISIBLE_DEVICES=0 python ConvGRU.py
python GRU.py
python LIM.py
python LR.py
3.1 Download the data to the dir ConvGRU_SPEAR/data
, and instructions can be found at ConvGRU_SPEAR/data/README.md
3.2 Train the ConvGRU network under the dir ConvGRU_SPEAR
, and the trained models will be saved in ConvGRU_CCSM4/results/saved_ConvGRU_model
CUDA_VISIBLE_DEVICES=0 python main.py
CUDA_VISIBLE_DEVICES=0 python ConvGRU.py
python GRU.py
python LIM.py
python LR.py
4.1 Download the data to the dir AirTmp_M/data
, and instructions can be found at AirTmp_M/data/README.md
4.2 Train the ConvGRU network under the dir AirTmp_M
, and the trained models will be saved in AirTmp_M/saved_ConvGRU_model
CUDA_VISIBLE_DEVICES=0 python main.py
CUDA_VISIBLE_DEVICES=0 python plot_res.py