search

xi-jia / Fourier-Net

41 stars 2 forks source link

For the issue related to PyTorch and CUDA version mismatch encountered during result replication. #1

Closed Marceloxo closed 8 months ago

Marceloxo commented 8 months ago

Thank you for your excellent work. I wanted to ask in what version of CUDA and PyTorch your code was trained and tested. I've been attempting to replicate your results, but I'm encountering issues where the loss doesn't decrease, and the Dice scores on the validation set don't improve. I suspect it might be due to a mismatch in PyTorch versions.

xi-jia commented 8 months ago

Thanks. The models were trained and tested on 2 different machines with details below.

> python
Python 3.8.12 (default, Oct 12 2021, 13:49:34) 
[GCC 7.5.0] :: Anaconda, Inc. on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import torch
>>> torch.__version__
'1.10.0+cu113'
$ python
Python 3.6.13 |Anaconda, Inc.| (default, Jun  4 2021, 14:25:59)
[GCC 7.5.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import torch
>>> torch.__version__
'1.10.0+cu102'
Marceloxo commented 8 months ago

Thank you for your prompt response. For experiments with the 'Fourier-Net' model, do you have a preference for using any of the mentioned environments? Thanks once again!

xi-jia commented 8 months ago

Not really, both the two envs produce similar results, as the Pytorch versions are the same.