Describe the bug
The bug is when i run the code it show's me the kernel is dead and restarting the main issue i found is the env method i created and running any gym method like env.render or env.action_space.sample() or env.close(). After I comment all this code or line then the code running fine but after running it without comment kernel will dead.
Code
import gym
from stable_baselines3 import PPO
from stable_baselines3.common.vec_env import DummyVecEnv
from stable_baselines3.common.evaluation import evaluate_policy
from IPython.display import display, clear_output
enviroment_name = 'CartPole-v1'
env = gym.make(enviroment_name, render_mode = 'human')
episodes = 5
for episodes in range(1, episodes + 1):
state = env.reset()
done = False
score = 0
while not done:
env.render()
action = env.action_space.sample()
n_state, reward, truncate, done, info = env.step(action)
score += reward
print('Episode:{} Score:{}'.format(episodes,score))
env.close()
Also console showing this:
KernelRestarter: restarting kernel (1/5), keep random ports
WARNING:root:kernel a23dd0d8-4dd7-4aac-a008-8736e29ab8c1 restarted
0.00s - Debugger warning: It seems that frozen modules are being used, which may
0.00s - make the debugger miss breakpoints. Please pass -Xfrozen_modules=off
0.00s - to python to disable frozen modules.
0.00s - Note: Debugging will proceed. Set PYDEVD_DISABLE_FILE_VALIDATION=1 to disable this validation.
System Info
i installed gym as pip install gym
Describe the characteristic of your environment:
i my enviroment i installed tensorflow and pytorch and most of the other things also cude is enabled in my enviroment.
Describe how Gym was installed (pip, docker, source, ...)
pip
What OS/version of Linux you're using. Note that while we will accept PRs to improve Window's support, we do not officially support it.
I am using linux mint
Additional context
I have checked the issue but still my problem isn't solved.
Checklist
[x] I have checked that there is no similar issue in the repo (required)
Describe the bug The bug is when i run the code it show's me the kernel is dead and restarting the main issue i found is the env method i created and running any gym method like env.render or env.action_space.sample() or env.close(). After I comment all this code or line then the code running fine but after running it without comment kernel will dead.
Code
Also console showing this:
System Info i installed gym as pip install gym Describe the characteristic of your environment: i my enviroment i installed tensorflow and pytorch and most of the other things also cude is enabled in my enviroment.
Additional context I have checked the issue but still my problem isn't solved.
Checklist