Replies: 2 comments 1 reply
-
Hi @edofazza, This is not a currently available feature in SLEAP -- multi-GPU support is a little tricky and OS-dependent. You're welcome to give it a go by following this guide on TensorFlow multi-device strategies together with this guide on training interactively with SLEAP. You'll want comment out the
You may need to tweak what goes inside the Cheers, Talmo |
Beta Was this translation helpful? Give feedback.
-
Hi @edofazza were you able to train with multiple gpu? |
Beta Was this translation helpful? Give feedback.
-
I have two GPUs and I want to use both for training, but the leap-train command takes as argument one GPU. Is there a way to use multiple GPUs at once to speed up the training?
Beta Was this translation helpful? Give feedback.
All reactions