fix pp memory issue (#6344)

Co-authored-by: Tong Li <tong.li35271158@gmail.com>
This commit is contained in:
Tong Li 2025-06-11 17:54:18 +08:00 committed by YeAnbang
parent 6ae54a6dce
commit 3a4681fdd9

View File

@ -358,7 +358,7 @@ class GRPOConsumer(BaseConsumer):
criterion=_criterion,
optimizer=self.optimizer,
return_loss=True,
return_outputs=True,
return_outputs=False,
)
loss = policy_model_outputs["loss"]