ethics-benchmark How to run python train.py You can change arguments(model_name, epochs, learning_rate, etc.) Reference Original Github