gym
gym copied to clipboard
NotAnyMike
An improvement of CarRacing-v0 from OpenAI Gym in order to make the environment complex enough for Hierarchical Reinforcement Learning
[Screenshot of the error](https://media.discordapp.net/attachments/832768230706905099/944597286812536852/Screenshot_from_2022-02-19_16-11-46.png?width=1178&height=662) My Friend was using the CarRacing-v1 and he's getting the keyerror and we cant seem to find any solution. More on that here [full issue](https://github.com/pyglet/pyglet/issues/546)
Hello, Was wondering if the model weights made available at https://notanymike.github.io/rl/2017/12/18/Solving-CarRacing.html were produced using the PPO hyperparameters from the original Schulman paper or did you tune for the CarRacing environment...
Sometimes a weird segment with very few tiles is created Before trying to solve this, given that this error is very rare, implement: - [x] save maps - [x] load...
- [ ] Mention Check_ouside - [ ] Mention Check_timeout - [ ] Mention S shortcut in all envs to visualize the env - [ ] Mention tests and how...
Right now if the reward is greater than 1000 it is solved
reuse_map in case of reset, reset everything, keep the map and start in a random location
Metadata
Owner
Metadata
An improvement of CarRacing-v0 from OpenAI Gym in order to make the environment complex enough for Hierarchical Reinforcement Learning