The code of "TC-Diffuser: Bi-Condition Multi-Modal Diffusion for Tropical Cyclone Forecasting" accepted by AAAI2025.

• python 3.8.8
• Pytorch 1.11.0 (GPU)

First, we need to download all the data we used in TC-Diffuser.

• TC-Diffuser's processed datasets
• TC-Diffuser's checkpoint

After completing the downloading, move these file to correct file path.

• Move TC-Diffuser's processed datasets to /TC-Diffuser, change the data_dir in TC-Diffuser/configs/baseline.yaml to TC-Diffuser/processed_data_noise_traj_inten_wind_gph_env
• Move TC-Diffuser's checkpoint to /TC-Diffuser/experiment

## change the eval_mode in TC-Diffuser/configs/baseline.yaml to False ##
cd TC-Diffuser
python main.py

## change the eval_mode in TC-Diffuser/configs/baseline.yaml to True ##
cd TC-Diffuser
python main.py

Part of our code is borrowed from MID. We thank the authors for releasing their code and models.