Hello, can I ask you some questions about your code?

Question

Hello, can I ask you some questions about your code?

Opened this issue 3 years ago · 2 comments

Excuse me, I have some questions:
First, I see that you are using PyTorch and what version of PyTorch framework you are using.
Second, compared with the program of DQN, does this DDPG use different versions of various feature packs?
That's all for now.
I really want to implement your code, and I am looking forward to your reply!

Answer 1 · 2021-07-01T03:31:18.000Z

In this project I used the Pytorch version 1.4.0, for your second question I didn't understand what are you asking. The DDPG is an Deep-RL technique for continuous control space and action, unlike DQN which works for continuous spaces but only has discrete actions.

Answer 2 · 2021-07-01T06:47:26.000Z

Thank you very much for your reply. My second question meant that I was using a Kenitic, Ubuntu-16.04, TensorFlow-CPU, etc., when I reproduced the DQN project for Turtlebot3. I would like to ask whether your project is still based on such an environment. And the GPU or CPU used by the PyTorch. When I was working on projects like this, I was often overwhelmed by environmental issues and version issues. Ubuntu has been reinstalled four times recently, and it's just terrible. Looking forward to your reply!Thank you very much for your reply. My second question meant that I was using a Kenitic, Ubuntu-16.04, TensorFlow-CPU, etc., when I reproduced the DQN project for Turtlebot3. I would like to ask whether your project is still based on such an environment. And the GPU or CPU used by the PyTorch. When I was working on projects like this, I was often overwhelmed by environmental issues and version issues. Ubuntu has been reinstalled four times recently, and it's just terrible. Looking forward to your reply! On 7/1/2021 11:31，Dranaju (Junior Costa de ***@***.***> wrote： In this project I used the Pytorch version 1.4.0, for your second question I didn't understand what are you asking. The DDPG is an Deep-RL technique for continuous control space and action, unlike DQN which works for continuous spaces but only has discrete actions. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.