search

reiniscimurs / GDAE

A goal-driven autonomous exploration through deep reinforcement learning (ICRA 2022) system that combines reactive and planned robot navigation in unknown environments
110 stars 15 forks source link

use pytorch and instead use TensorFlow #17

Open YMUDRL opened 3 months ago

YMUDRL commented 3 months ago

Hi Reinis Cimurs, May I ask why not continue to use pytorch and instead use TensorFlow here? Is it feasible for me to change it to versions after TensorFlow 2.0? It seems that version 1.14 cannot be downloaded now

reiniscimurs commented 3 months ago

Hi,

This repository is already some 4 years old and as such the technology used here may be outdated. The DRL navigation repository and the code was created after this work was done so different code base is used there. In theory, the env file does not care what is used - pytorch or tensorflow. All it needs is an input of an action. So pytorch model from the DRL repository could also be used here, given that state is returned accordingly. Feel free to update the necessary code parts and there are no hard requirements for the model part.

YMUDRL commented 3 months ago

Thank you so much, this way I don't have to change my code to adapt to tensorflow 2.0, and continue to use my pytorch! By the way, is there any progress in your work based on GDAE, and are you considering expanding it to UAVs or multi-intelligence, your work has inspired me so much, thank you again from the bottom of my heart!

reiniscimurs commented 3 months ago

Thank you so much, this way I don't have to change my code to adapt to tensorflow 2.0, and continue to use my pytorch! By the way, is there any progress in your work based on GDAE, and are you considering expanding it to UAVs or multi-intelligence, your work has inspired me so much, thank you again from the bottom of my heart!

This was part of my PhD work and I have since moved on to a different position in a slightly different field. There are no plans to work on this in the future.