search

NVlabs / SegFormer

Official PyTorch implementation of SegFormer
https://arxiv.org/abs/2105.15203
Other
2.36k stars 332 forks source link

Issue about setting number of GPUs for evaluation during training process #133

Closed Yonghao-Tan closed 9 months ago

Yonghao-Tan commented 9 months ago

I can adjust the number of GPUs in direct test. However, I found the evaluation during the training process quite slow due to a single GPU. Where can I set the number of GPUs for evaluation during the training process? Thanks!