gym
gym copied to clipboard
A toolkit for developing and comparing reinforcement learning algorithms.
### Question Do you happen to know if state signals in Classic Control, Box2D & MuJoCo tasks all possess the Markov Property?
Hi, I am trying to define a observation space in my custom env that, I have an agent flying in a 3-d space, and there are serveral obstacles in the...
### Question My gym version is `0.17.0`. I create an `Hopper-v2` environment. When I set seed of 10 using `env.action_space.seed(seed)`, I get the following output: ``` env.action_space.seed(10) [10] env.action_space.sample() array([...
#### In the [pendulum.py file](https://github.com/openai/gym/blob/master/gym/envs/classic_control/pendulum.py), a file named pendulum.png explaining the coordinate system is said to be located at './diagrams/pendulum.png' in the description comments of PendulumEnv class. I am not...
### Proposal Get the is_slippery parameter infos more visible in the documentation ### Motivation When running Frozen Lake for the fisrt time, is_slippery add a default parameter set at true....
### Question Hi all, I am now using the CarRacing-v2 to train my RL algorithm. when I make the environment by: ``` env = gym.make("CarRacing-v2").unwrapped ``` I found that I...
When setting up my envs, I get Value Error messages for my env.step(env.action_space.sample()) If you can help that will be greatly appreciated! (Mac OS, Jupyter Lab) Here's the Code: !pip3...
**Describe the bug** Before gym 0.26: https://github.com/openai/gym/blob/3498617bf031538a808b75b932f4ed2c11896a3e/gym/wrappers/time_limit.py#L51-L56 After: https://github.com/openai/gym/blob/54b406b7991e49ab917ed355d12971009e8248b7/gym/wrappers/time_limit.py#L53-L54 It should be `truncated = truncated or not terminated` to match previous behavior. test was also updated in https://github.com/openai/gym/commit/54b406b7991e49ab917ed355d12971009e8248b7#diff-15620e7669a4869759c87e2b6a4f51ad87609ed2680a54047e2da2ff7fcc4149 If the...
**Description** If the `seed` argument you pass to `env.reset(seed=seed)` is not an integer, the raised error is uninformative as it only includes the seed *value* not the seed *type*. **Code...
### Question Hi, I have a question on the state of Blackjack-v1. In Blackjack-v1: 1. env.observation_space = Tuple(Discrete(32), Discrete(11), Discrete(2)) 2. However, I got a state `(**47**, 8, False)`. **The...