You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hello, i have this problem where i at some point start running out of memory when running python train.py --threed_match_dir ~/dataset/threedmatch/ --batch_size 1.
At first i tried with batch_size 2 but it was too much for my gpu so i changed it to 1. After going through some thousands epochs i started getting "out of memory" errors like:
INFO - 2021-02-22 12:51:28,348 - trainer - Train Epoch: 1 [1440/7317], Current Loss: 1.157e+00 Pos: 0.365 Neg: 0.792 Data time: 0.0536, Train time: 0.5614, Iter time: 0.6150
Traceback (most recent call last):
File "train.py", line 84, in <module>
main(config)
File "train.py", line 63, in main
trainer.train()
File "/home/f/repos/FCGF/lib/trainer.py", line 132, in train
self._train_epoch(epoch)
File "/home/f/repos/FCGF/lib/trainer.py", line 492, in _train_epoch
self.config.batch_size)
File "/home/f/repos/FCGF/lib/trainer.py", line 427, in contrastive_hardest_negative_loss
D01 = pdist(posF0, subF1, dist_type='L2')
File "/home/f/repos/FCGF/lib/metrics.py", line 24, in pdist
D2 = torch.sum((A.unsqueeze(1) - B.unsqueeze(0)).pow(2), 2)
RuntimeError: CUDA out of memory. Tried to allocate 32.00 MiB (GPU 0; 3.82 GiB total capacity; 744.27 MiB already allocated; 43.38 MiB free; 814.00 MiB reserved in total by PyTorch)
Currently my system takes up 500MiB VRAM from my GTX 1650 (4GB) and the rest is used by pytorch. I'm running pytorch 1.7 in a python 3.7 conda enviroment and i tried running tried compiling minkowskiEngine for cuda 11.2 and 10.2 but both gave the same error.
The text was updated successfully, but these errors were encountered:
Hello, i have this problem where i at some point start running out of memory when running
python train.py --threed_match_dir ~/dataset/threedmatch/ --batch_size 1
.At first i tried with batch_size 2 but it was too much for my gpu so i changed it to 1. After going through some thousands epochs i started getting "out of memory" errors like:
Currently my system takes up 500MiB VRAM from my GTX 1650 (4GB) and the rest is used by pytorch. I'm running pytorch 1.7 in a python 3.7 conda enviroment and i tried running tried compiling minkowskiEngine for cuda 11.2 and 10.2 but both gave the same error.
The text was updated successfully, but these errors were encountered: