EMAWeightAveraging updates the average model on steps skipped by the AMP GradScaler#21990 · open · 0 comments
Add overloads to `Fabric.setup`, `setup_optimizers`, and `setup_dataloaders` for precise return types#21988 · open · 0 comments
Mixed precision uses `torch.autocast("cuda")` for custom accelerators and `Accelerator` instances, silently running fp32#21982 · open · 0 comments
RichProgressBar: the refresh at on_validation_epoch_end renders the previous epoch's metrics#21972 · closed · 1 comments