search

Vincent-ZHQ / MRDF

Code for Cross-Modality and Within-Modality Regularization for Audio-Visual DeepFake Detection
Apache License 2.0
18 stars 2 forks source link

The amount of GPU usage during training #3

Closed guishengzhanga closed 5 months ago

guishengzhanga commented 6 months ago

Hello, I'm interested in your work. Can you tell me on which type of graphics card the training was conducted and how much GPU usage there is for each card?

Vincent-ZHQ commented 5 months ago

I remember I used one GPU of A100 (40g) for training. The default batch size is 64. So I think it could run in any GPU when you adapt the batch size.