fix formatting issue

Signed-off-by: jouw <jouw@foxmail.com>

Author: jouw

applications/DeepSpeed-Chat/dschat/rlhf/ppo_trainer.py

@@ -236,7 +236,8 @@ def train_rlhf(self, inputs):
         value = self.critic_model.forward_value(**batch,
                                                 return_value_only=True,
                                                 use_cache=False)[:, :-1]
-        critic_loss = self.critic_loss_fn(value[:, start:], old_values[:, start:],
+        critic_loss = self.critic_loss_fn(value[:, start:], old_values[:,
+                                                                       start:],
                                           returns, action_mask[:, start:])
         self.critic_model.backward(critic_loss)
 

applications/DeepSpeed-Chat/training/step3_rlhf_finetuning/main.py

@@ -594,7 +594,8 @@ def main():
                     "-------------------------------------------------------------------------------------",
                     args.global_rank)
 
-                if args.enable_tensorboard and torch.distributed.get_rank() == 0:
+                if args.enable_tensorboard and torch.distributed.get_rank(
+                ) == 0:
                     writer.add_scalar('reward',
                                       average_reward / inner_iter,
                                       global_step=step)