Hello. I'm just student.
I'm not good at English so Excuse me first.
In https://github.com/yandexdataschool/Practical_RL/blob/master/week04_approx_rl/homework_pytorch_main.ipynb, class DQNAgent's forward uses n_actions of global variables as following.
"assert len(
qvalues.shape) == 2 and qvalues.shape[0] == state_t.shape[0] and qvalues.shape[1] == n_actions"
I guess assert should use class private variables
assert len(
qvalues.shape) == 2 and qvalues.shape[0] == state_t.shape[0] and qvalues.shape[1] == **self.**n_actions
It's not important I know.
Just want to check whether my thought is ok...
Thank you!
If you need, tell me my e-mail : 6129876j@naver.com.
Hello. I'm just student.
I'm not good at English so Excuse me first.
In https://github.com/yandexdataschool/Practical_RL/blob/master/week04_approx_rl/homework_pytorch_main.ipynb, class DQNAgent's forward uses n_actions of global variables as following.
"assert len(
qvalues.shape) == 2 and qvalues.shape[0] == state_t.shape[0] and qvalues.shape[1] == n_actions"
I guess assert should use class private variables
assert len(
qvalues.shape) == 2 and qvalues.shape[0] == state_t.shape[0] and qvalues.shape[1] == **self.**n_actions
It's not important I know.
Just want to check whether my thought is ok...
Thank you!
If you need, tell me my e-mail : 6129876j@naver.com.