Noisy_DQN

Atari-Noisy-DQN

Train NoisyNet DQN

$ python main.py --train --save_path <model_path> --noisy

Resume Training

$ python main.py --train --resume --resume_path <previous_saved_model_path> --save_path <new_model_path> --best <previous_best_ep_score> --noisy

Test model

$ python main.py --test --save_path <saved_model_path> --noisy

Result