close
The Wayback Machine - https://web.archive.org/web/20220329033044/https://github.com/topics/openai-gym
Skip to content
#

openai-gym

Here are 959 public repositories matching this topic...

scaro86
scaro86 commented Mar 21, 2022

Hi,
I am currently using your FinRL_PortfolioAllocation_NeurIPS_2020 code and I have some strange behavior at the beginning of training. Sometimes the first episode reward mean value is super high and then drops during the training as shown on the tensorboard plot. This high value is never reached again. Any idea why this is happening ?

Edit: I'm training PPO agent from stablebaselines3 wit

bug good first issue
gym-super-mario-bros
Kautenja
Kautenja commented Jan 4, 2019

Is your feature request related to a problem? Please describe.

No

Describe the solution you'd like

A render mode that utilizes the x y position of Mario and static maps of each level to plot a simple point on a graph. Each episode could overlay as an "extended exposure" of the training, validation, or testing procedure.

Describe alternatives you've considered

NA

A

enhancement help wanted good first issue

Improve this page

Add a description, image, and links to the openai-gym topic page so that developers can more easily learn about it.

Curate this topic

Add this topic to your repo

To associate your repository with the openai-gym topic, visit your repo's landing page and select "manage topics."

Learn more