search

hkchengrex / Cutie

[CVPR 2024 Highlight] Putting the Object Back Into Video Object Segmentation
https://hkchengrex.com/Cutie/
MIT License
732 stars 71 forks source link

Model training using GPUs on multiple computers #109

Closed panyuxin1993 closed 2 months ago

panyuxin1993 commented 2 months ago

I like the project a lot and want to train model using our own data. However, we do not a powerful computer with so many powerful GPUs. I am wondering whether it is possible to run the training using multiple computers, i.e. a cluster?

hkchengrex commented 2 months ago

You can see https://pytorch.org/tutorials/beginner/dist_overview.html for tips. I have not tried multi-node parallel training.