OpenAI Gym is a platform for reinforcement learning research that aims to provide a general-intelligence benchmark with a wide variety of environments.
Questions tagged [openai-gym]
1033 questions
0
votes
1 answer
tflearn loss is always 0.0 while training reinforcement learning agent
I tried to train a reinforcement learning agent with gym and tflearn using this code:
from tflearn import *
import gym
import numpy as np
env = gym.make('CartPole-v0')
x = []
y = []
max_reward = 0
for i in range(1000):
env.reset()
while…
Kay Jersch
- 277
- 3
- 13
0
votes
1 answer
How does Deep Q learning work
When I am training my model I have the following segment:
s_t_batch, a_batch, y_batch = train_data(minibatch, model2)
# perform gradient step
loss.append(model.train_on_batch([s_t_batch, a_batch], y_batch))
where s_t, a_ corresponds to current…
sachinruk
- 9,571
- 12
- 55
- 86
0
votes
1 answer
Policy gradient methods for Open AI Gym Cartpole
I am a beginner in Reinforcement Learning and am trying to implement policy gradient methods to solve the Open AI Gym CartPole task using Tensorflow. However, my code seems to run extremely slowly; the first episode runs at an acceptable pace,…
ashboy64
- 83
- 3
- 13
0
votes
1 answer
OpenAI can't get program to start
Whenever I try to run the program on universe, I get this error where the sockets close. I don't know how to fix it and I have been messing around with it for ages. I am using the sample code given by openAI, but for some reason the flash game…
memememe
- 1
- 1
0
votes
0 answers
Implementing Q Learning For OpenAi CartPole in Python (v0)
--------------------------------------------------------------UPDATE---------------------------------------------------------
I continued trying to improve the code by altering the parameters and the way it updates through the learning rule. I…
ashboy64
- 83
- 3
- 13
0
votes
1 answer
Function approximator and q-learning
I am trying to implement q-learning with an action-value approximation-function. I am using openai-gym and the "MountainCar-v0" enviroment to test my algorithm out. My problem is, it does not converge or find the goal at all.
Basically the…
mrmjauh
- 71
- 1
- 9
0
votes
1 answer
Cannot disable GUI for FCEUX 2.2.2
I am trying to run the FCEUX emulator on AWS without graphical rendering so that my machine learning algorithm can train faster. So far I have tried running FCEUX from command line with the following commands:
fceux --nogui 1…
user3543300
- 499
- 2
- 9
- 27
0
votes
1 answer
OpenAI gym keyboard_agent.py: How to restart?
I want to use the keyboard_agent.py to play Breakout-v0.
When I start the environment with:
python [...]/gym/examples/agents/keyboard_agent.py "Breakout-v0"
The game starts, and I can play the first round no problem. I can't however make it…
Toke Faurby
- 5,788
- 9
- 41
- 62
0
votes
1 answer
Deep Neural Network does not update weights upon training
I am currently getting into tensorflow and have just now started to grasp the graph like concept of it. Now I tried to implement a NN using gradient descent(Adam optimizer) to solve the cartpole environment. I start by randomly intializing my…
jaaq
- 1,188
- 11
- 29
0
votes
1 answer
Python OpenAI Atari module in Ubuntu 16.04 - "No module named atari_py."
(I'm new to Ubuntu, excuse any ignorance.) I'm trying to run SpaceInvaders in Python using OpenAI's Atari module, however I keep getting the error
File
"/usr/local/lib/python2.7/dist-packages/gym/envs/atari/atari_env.py",
line 11, in raise…
revvy
- 1
- 2
0
votes
0 answers
Ubuntu on Windows running openAI gym, env.render() give out nothing
import gym
env = gym.make('Pong-v0')
env.reset()
for _ in range(10):
env.render(mode='rgb_array')
# I tried env.render(close = True) and env.render(close = False)
env.step(env.action_space.sample()) # take a random action
# env =…
user6924524
- 1
- 1
0
votes
1 answer
How can I use other environments
OpenAI's universe is a awesome library, Since the code
# coding: utf-8
import gym
import universe # register the universe environments
env = gym.make('flashgames.DuskDrive-v0')
env.configure(remotes=1) # automatically creates a local docker…
GoingMyWay
- 16,802
- 32
- 96
- 149
0
votes
1 answer
Openai universe-starter-agent not training
I've been trying to run Openai's universe-starter-agent example found here, However, using an m4.16xlarge instance on AWS with 32 workers, the agent's training result doesn't improve after 0.6 hours (over 30 minutes) while it is stated that "the…
Charmander
- 109
- 14
0
votes
1 answer
OpenAI "gym" module cannot be imported with PyCall in Julia
I try to import gym module of python inside julia by using PyCall.jl library.
To make it, I applied the followings:
$ julia
julia> using PyCall
julia> gym = pyimport("gym")
ERROR: PyError (:PyImport_ImportModule)
zwlayer
- 1,752
- 1
- 18
- 41
0
votes
1 answer
'VNCEnv' object has no attribute 'remote_manager' when running OpenAI universe example
When I try to run the following example code from here, Python is unable to find the remote_manager attribute. I have all required packages installed and the OpenAI gym examples appear to work as they should.
I run Python 3.5 under Linux Mint 18,…
wehnsdaefflae
- 826
- 2
- 12
- 27