Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Can't reproduce DQN performance #14

Closed
SunCherry opened this issue Oct 18, 2019 · 2 comments
Closed

Can't reproduce DQN performance #14

SunCherry opened this issue Oct 18, 2019 · 2 comments

Comments

@SunCherry
Copy link

I noticed you changed the optimizer and some hyper-parameters in DQN compared to those in the "Nature" paper, well, from my side I can't reproduce results by taking any of the two settings, could you share a learning curve of "Breakout"? I have been struggling with the hyper-parameters optimization for two months. Thanks.

@aslanides
Copy link
Member

Hi there.

This agent is intended to be a simple instantiation of the DQN algorithm (Q-learning + non-linear function approximation + experience replay), and isn't intended to reproduce the Nature Atari results. There are numerous subtleties related to interfacing with Atari (frame stacking, reward clipping, etc) that can be tricky to get right. For agents that are set up to run on Atari, see Dopamine (github.com/google/dopamine) or OpenAI baselines (github.com/openai/baselines)

@SunCherry
Copy link
Author

Hi there.

This agent is intended to be a simple instantiation of the DQN algorithm (Q-learning + non-linear function approximation + experience replay), and isn't intended to reproduce the Nature Atari results. There are numerous subtleties related to interfacing with Atari (frame stacking, reward clipping, etc) that can be tricky to get right. For agents that are set up to run on Atari, see Dopamine (github.com/google/dopamine) or OpenAI baselines (github.com/openai/baselines)

Got it, thanks a lot.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants