search

drprojects / superpoint_transformer

Official PyTorch implementation of Superpoint Transformer introduced in [ICCV'23] "Efficient 3D Semantic Segmentation with Superpoint Transformer" and SuperCluster introduced in [3DV'24 Oral] "Scalable 3D Panoptic Segmentation As Superpoint Graph Clustering"
MIT License
560 stars 72 forks source link

how to calculate the preprocessing time #151

Closed yeeinn closed 4 weeks ago

yeeinn commented 1 month ago

How to calculate the time spent on preprocessing. it is 12.4min of s3dis in your paper .

drprojects commented 4 weeks ago

The preprocessing times for the train, val, and test set all appear in the logs (see progress bars) when instantiating a dataset for the first time and preprocessing is triggered.

Otherwise, you can time whatever piece of code you are interested in with:

import torch
from time import time

torch.cuda.synchronize()
start = time()

# stuff happens...

torch.cuda.synchronize()
end = time()

print(end - start)