search

alexfrom0815 / Online-3D-BPP-DRL

This repository contains the implementation of paper Online 3D Bin Packing with Constrained Deep Reinforcement Learning.
292 stars 66 forks source link

the training speed problem #11

Closed cakaex closed 2 years ago

cakaex commented 2 years ago

when the bin size and box size increase, the training is very slow. is there any way to improve it?

alexfrom0815 commented 2 years ago

when the bin size

Hi, when the bin size increases, the original heightmap input (state) to the CNN also gets larger, which slows down the algorithm.

The good news is that our new work no longer uses the heightmap representation, the input container size and item size can be arbitrary and this does not affect the speed of training. If you are interested, you can find our new job at https://github.com/alexfrom0815/Online-3D-BPP-PCT.