Explain general purpose of the work...
Explain generate_synthetic_data.py and the structure of the datasets.
Explain download_data.{py,sh} and the downloadable datasets.
The teacher setting ...
python3 run_teacher.py --dataset_filename ) [DATASET_FILENAME]
--initial_relabeling ) {'ones', 'degrees'}
# -v | --verbose ) [VERBOSE]
# --save_file_destination ) [SAVE_FILE_DESTINATION]
--num_iterations ) [NUM_ITERATIONS]
--setting ) {'regression', 'classification'}
--classes ) [CLASSES]
--bias ) [BIAS]
--lower_bound ) [LOWER_BOUND]
--upper_bound ) [UPPER_BOUND]
{'GIN'} ...Commented arguments are mostly used for internal development. After all the optional arguments, one must specify the GNN model as a positional argument, followed by all its specific arguments detailed in previous sections. For more information on the arguments, types and descriptions, run python3 run_teacher.py --help.
The student setting ...
python3 run_model.py --
TBD
TBD