remove useless code

Author: wangbluo

colossalai/shardformer/modeling/mistral.py

@@ -270,7 +270,6 @@ def mistral_for_causal_lm_forward(
                 shift_labels = labels[..., 1:].contiguous()
                 # Flatten the tokens
                 loss_fct = CrossEntropyLoss()
-                #shift_logits = shift_logits.view(-1, self.config.vocab_size)
                 shift_labels = shift_labels.view(-1)
                 # Enable model parallelism
                 shift_labels = shift_labels.to(shift_logits.device)

colossalai/shardformer/policies/mistral.py

@@ -277,7 +277,6 @@ def module_policy(self):
                             suffix="lm_head",
                             target_module=VocabParallelLMHead1D,
                             kwargs={
-                                #gather_output=True,
                                 "gather_output": not self.shard_config.parallel_output,
                                 "make_vocab_size_divisible_by": self.shard_config.make_vocab_size_divisible_by,
                             },