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
Results object reduction when batch sizes are different won't work because torch.stack get's different input shapes. This can happen if your dataloader returns a smaller batch for the last iteration, for example.
awaelchli
changed the title
Results reduction when batch size changes over time (e.g. last batch)
Results gathering with varying tensor shapes (e.g. last batch)
Aug 17, 2020
🐛 Bug
Results object reduction when batch sizes are different won't work because torch.stack get's different input shapes. This can happen if your dataloader returns a smaller batch for the last iteration, for example.
Context
From slack discussion by @artgor
https://pytorch-lightning.slack.com/archives/CRBLFHY79/p1597604494424600
The text was updated successfully, but these errors were encountered: