search

jeya-maria-jose / Medical-Transformer

Official Pytorch Code for "Medical Transformer: Gated Axial-Attention for Medical Image Segmentation" - MICCAI 2021
MIT License
791 stars 176 forks source link

Question about the training speed #52

Open JackHeroFly opened 2 years ago

JackHeroFly commented 2 years ago

Hello, author. Thank you for your code. CUDA has been used and the batch size has been adjusted to 4, but the training speed of the model is still very slow. Are you in this situation? How can I solve this problem? Look forward to your reply.

MukulKadaskar commented 1 year ago

Hello, author. Thank you for your code. CUDA has been used and the batch size has been adjusted to 4, but the training speed of the model is still very slow. Are you in this situation? How can I solve this problem? Look forward to your reply.

I am also facing the same problem.

MukulKadaskar commented 1 year ago

Even though having multiple GPUs, its still not utilising GPU to full extent and training very slowly.

canglangzhige commented 8 months ago

Hello, author. Thank you for your code. CUDA has been used and the batch size has been adjusted to 4, but the training speed of the model is still very slow. Are you in this situation? How can I solve this problem? Look forward to your reply.

Hi, I am facing the same problem. The reason is that the code uses the "for" in Python when processing the local branch. But I don't have any method to solve it.