How do I use GPU to accelerate training?

[JieJayCao]

Hello, thank you for reproducing mlp-mixer, but in the process of using and training my own datasets, I found that the speed is slower and GPU acceleration is not used. How can I solve this problem? Looking forward to your answer.

[bangoc123]

You can simply reduce the argument batch-size to 4 or 2 to solve the problem. About the GPU, you run this repo on colab or your own machine? Vào Th 3, 30 thg 11, 2021 lúc 19:27 JieCao ***@***.***> đã viết:

…

Hello, thank you for reproducing mlp-mixer, but in the process of using and training my own datasets, I found that the speed is slower and GPU acceleration is not used. How can I solve this problem? Looking forward to your answer. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#2>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADG3ZHJCJ5JWZDQ6CS5LIWTUOS7JHANCNFSM5JBVTB7Q> .