Submission result differs from local

stanfordnmbl / osim-rl

Reinforcement learning environments with musculoskeletal models

http://osim-rl.stanford.edu/

MIT License

882 stars 249 forks source link

Closed whikwon closed 5 years ago

whikwon commented 5 years ago

I ran the evaluation code and could get 1999 reward. but when i submit the result, total reward becomes half value. Why this happens?

kidzik commented 5 years ago

Please follow these instructions https://github.com/stanfordnmbl/osim-rl#submission basically, you need to restart the environment and send the next set of actions for the second episode.

While in the first round the result might be the same, in the second the random seeds are different and the results from the runs will vary.

whikwon commented 5 years ago

Thanks