You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
GPU acceleration is important for improving model inference performance. GPU has its own memory and the DL framework like libtorch has its own memory management mechanism. Loading too many models on GPU will cause OOM issue. We should control memory usage to ensure high availability.
The text was updated successfully, but these errors were encountered:
GPU acceleration is important for improving model inference performance. GPU has its own memory and the DL framework like libtorch has its own memory management mechanism. Loading too many models on GPU will cause OOM issue. We should control memory usage to ensure high availability.
The text was updated successfully, but these errors were encountered: