云服务器运行持续强化学习COOM框架的问题

1 环境要求

下载地址：https://github.com/TTomilin/COOM
tensorflow 2.11以上
python 3.9以上
tensorflow2.12.0，需要安装tensorflow-probability0.19

2 修改代码

COOM/wrappers/reward.py
将

from gym import RewardWrapper

修改为

from gymnasium import RewardWrapper

3 云服务器运行程序

需要安装pip install opencv-python-headless4.8.1.78，兼容numpy1.23.5
否则运行测试程序会报错

from COOM.env.builder import make_env  
from COOM.utils.config import Scenario  env = make_env(Scenario.RAISE_THE_ROOF)  
env.reset()  
for steps in range(10):  action = env.action_space.sample()  state, reward, done, truncated, info = env.step(action)  # env.render()  # 注释掉这行  if done:  break  
env.close()
print("成功")

报错：

qt.qpa.plugin: Could not find the Qt platform plugin “offscreen” in “/usr/local/miniconda3/lib/python3.10/site-packages/cv2/qt/plugins” This application failed to start because no Qt platform plugin could be initialized. Reinstalling the application may fix this problem.

报错

如果报错
ModuleNotFoundError: No module named ‘cv2’
重新强制安装

# 卸载残留包
pip uninstall -y opencv-python opencv-python-headless# 重新安装指定版本
pip install opencv-python-headless==4.8.1.78 --force-reinstall# 验证安装
python -c "import cv2; print(cv2.__version__)"