search

JingyunLiang / RVRT

Recurrent Video Restoration Transformer with Guided Deformable Attention (NeurlPS2022, official repository)
https://arxiv.org/abs/2206.02146
Other
354 stars 33 forks source link

The RAM is instantly full #7

Open Capsulers opened 1 year ago

Capsulers commented 1 year ago

The RAM increases sharply during the running of the main_test_rvrt.py, causing the process to be killed. I tried to lower the tile value, but it still didn't work. I ran this model on a Tesla P100. I would like to ask you how to solve this problem.

mrgreywater commented 1 year ago

For me it needs around 50GB of RAM to process a sample, you may just need to increase your swap file if you haven't installed enough RAM.

MelihDarcanxyz commented 7 months ago

My RAM wasn't enough so I rented a cloud server with A100 (80GB of VRAM) and 250GB of RAM. Unfortunately, it allocated all 250GB of RAM and the process has crashed. Is this normal? Or is there a memory leak somewhere in the code? To this day, I had no luck getting this model to work so I wonder.