Unverified Commit 202e0bbe authored by Myle Ott's avatar Myle Ott Committed by GitHub
Browse files

Fix bug when training with FP32 and --update-freq (#236)

parent 9143dfab

fairseq/trainer.py

+0 −1
Changes for fairseq/trainer.py: 0 added lines, 1 removed line.
Original line number Diff line number Diff line
@@ -198,7 +198,6 @@ class Trainer(object): 
                self.model.eval()

            else:

                self.model.train()

                self.optimizer.zero_grad()



            if sample is not None:

                with torch.no_grad() if eval else contextlib.ExitStack():