search

Kaixhin / Rainbow

Rainbow: Combining Improvements in Deep Reinforcement Learning
MIT License
1.56k stars 282 forks source link

To run a demo #65

Closed zyzhang1130 closed 4 years ago

zyzhang1130 commented 4 years ago

Hi, may I ask after the training is how should I run the trained agent to perform the task in with its fixed parameters for demonstration of its performance?

Thank you.

Kaixhin commented 4 years ago

If you read the main file, you'll see adding the --evaluate argument on the command line will do this. This combines with --model <model_path> for loading a saved agent. This is how evaluation is done, instead of training the model.

zyzhang1130 commented 4 years ago

noted with thanks. I will try it out.