pde-observers

1. Uncomment gt and set collect_data=true to generate data.

python run_pde_observers.py --train_yaml configs/matlab_rno.yaml

2. Set data_folder to 'outputs/{exp_name}' to train rno and evaluate. The command is the same.

python run_pde_observers.py --train_yaml configs/matlab_rno.yaml

old instructions for reference.

1. First run the following to preprocess data (for efficient running and debugging):

python lib/mat2npy.py

2. Then run to train the supervised pde observers:

python run_pde_observers.py
# rno
python run_pde_observers.py --train_yaml configs/matlab_rno.yaml
python run_pde_observers.py --train_yaml configs/base_transformer.yaml

3. Run the controller via:

python run_control.py