远程环境无法渲染
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
| import gymnasium as gym
from stable_baselines3 import A2C
env = gym.make("CartPole-v1", render_mode="rgb_array")
model = A2C("MlpPolicy", env, verbose=1)
model.learn(total_timesteps=10_000)
vec_env = model.get_env()
obs = vec_env.reset()
for i in range(1000):
action, _state = model.predict(obs, deterministic=True)
obs, reward, done, info = vec_env.step(action)
vec_env.render("human")
# VecEnv resets automatically
# if done:
# obs = vec_env.reset()
|
会得到报错NoSuchDisplayException: Cannot connect to "None"，原因是远程服务器没有接受显示的媒体。那么解决的第一个方案就是通过一个虚拟桌面接受来使其不报错，然后通过逐帧绘制图像，将过程展示在画板上。
XVFB + Matplotlib
XVFB
Xvfb是一个实现了X11显示服务协议的显示服务器。
安装
运行代码
1
| xvfb-run -s \"-screen 0 1400x900x24\" python <your_script.py>
|
运行jupyter
1
| xvfb-run -s "-screen 0 1400x900x24" jupyter-notebook
|
通过matplotlib渲染过程
1
2
3
4
5
6
7
8
9
10
11
12
13
14
| from IPython import display
from matplotlib import pyplot as plt
%matplotlib inline # turn on in cell
env = gym.make(..., render_mode="rgb_array") # must set `render_mode` to "rgb_array"
...
img = plt.imshow(vec_env.render())
for i in range(n):
...
img.set_data(vec_env.render())
display.display(plt.gcf())
display.clear_output(wait=True)
|
示例代码
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
| import gymnasium as gym
from stable_baselines3 import A2C
from IPython import display
from matplotlib import pyplot as plt
%matplotlib inline
env = gym.make("CartPole-v1", render_mode="rgb_array")
model = A2C("MlpPolicy", env, verbose=1)
model.learn(total_timesteps=10_000)
vec_env = model.get_env()
obs = vec_env.reset()
img = plt.imshow(vec_env.render())
for i in range(1000):
action, _state = model.predict(obs, deterministic=True)
obs, reward, done, info = vec_env.step(action)
img.set_data(vec_env.render())
display.display(plt.gcf())
display.clear_output(wait=True)
# VecEnv resets automatically
# if done:
# obs = vec_env.reset()
|
其他方案
参考 StackOverflow
