Where communities thrive

  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
    Jack C Crawford
    This message was deleted
    Ian Goodfellow
    ^ Channel moved.
    Just a friendly reminder that this channel is closed and moved
    This message was deleted
    Ian Goodfellow
    Just a friendly reminder that this channel is closed and moved
    Room dead?
    Pooja Gadige
    I can't seen to access the room https://gitter.im/openai/socml16 I see an error that says, "404. this is not the room you're looking for"
    is the room accessible only to some?
    I can't access it either
    Hi, is there anything wrong with this project? I am trying to find its api doc, and can't find anything
    can anyone tell me where is the project's api doc?
    There doesn't seem to be much of an api doc
    From what I could gather, OpenAI primarily uses gym and universe internally, but the wrappers were released to the public for some reason
    Without too much documentation
    Ozzie Zhang
    why openai gym chat room is not exist now?
    This chat room is still here, but the openai/socml16 chat room doesn't seem to exist. @goodfeli is there a new chat room or is this now the official openai room again?
    Hello how to get a api-key now , i forget the key i've applied before! and where to view the score now?
    Hello, i am trying to make an agent for the game "tennis-v0", my data is the raw image. I would like to know some techinque to locate players positions, ball position, and score in order to extract those features. Can anyone help refering me to some resource? Thanks
    Hello, I have looked at the Cartpole problem in gym. I have solved. Can anyone tell me what the best perfomance in the shortest episodes was?
    Or what the benchmark of the cartpole v0 was. I really need it to check how well my algorithm performed
    the title shows this room is moved to openai/socml16, but i didn't find this room. Is there any links?
    As everyone else the room https://gitter.im/openai/socml16 is not available, plus I can't find the gym chatroom as well.
    Ankit Choudhary
    Hi, I wish to render and record a gif from Taxi-V2 environment. The problem is that env.render does not output an image for this environment. Can someone suggest an alternative for the same?
    Eugene Valassakis
    Hello, has anyone worked out how to change mujoco model parameters on the fly (such as body_mass or geom_friction)? Threads openai/gym#234 and openai/gym#1262 don't seem to have an answer yet apart from changing the xml and reloading!
    Rodolfo Mendes
    hello everybody
    hello everygirle
    Chris Strobl
    hi there!
    Nicola Russo

    Hi all!! I'm a software engineer starting playing with OpenAI and robotics :) I'm looking for a genuine feedback/thoughts on my last project: https://github.com/nicrusso7/rex-gym

    this is a 3d printed quadruped and a 'run' gym env. thanks!

    Galen Wilkerson
    Hi there, I'm very new here, I've noticed there are many cool environments on gym.openai.com. I'm wondering if there is a simple pre-existing 2-D stick figure available, and how to get started with the basics on gym. thanks for any help!
    anyone here?
    Franklin Shih
    Hi, I am new here and have a few generic questions. any help is appreciated :)
    is anyone know how to get more information about gym's environment?
    I guess this room is dead, so I am not sure why you still keep it around
    Hello. Anyone got any idea how to approach an Optimal Control task, where the control matrices vary (slightly ie B'=0.8B, and only one matrix changes at a time) every time the environment is reset? Currently using DDPG with the state being a trajectory. Thanks for the help!!
    Franklin Shih
    may i know the diifferences between ray train and tune?
    what i mean is can ray tune converge?
    ray tune is slow compared to train so how about i train the model first and then do the hyperparameter tuning?
    Hi. Could someone please verify this issue? I am looking to know how pendulum-v1 env dynamics equations are derived. I am working on a project with MPC and a DDPG agent. A properly tuned MPC doesnt do well on the dynamics provided in gym. I may be wrong and I am stuck for a long time unable to verify how correct the gym dynamics are. It would be incredible if someone can look into this issue. Thanks IA. openai/gym#2999