From d38e6f954ec29b96fe2cf033937dad64e279b5d9 Mon Sep 17 00:00:00 2001 From: Gustavo de Rosa Date: Thu, 26 Oct 2023 20:01:15 +0000 Subject: [PATCH] Update modeling_mixformer_sequential.py Removes print regarding attention_mask to prevent excessive information from being logged. --- modeling_mixformer_sequential.py | 3 --- 1 file changed, 3 deletions(-) diff --git a/modeling_mixformer_sequential.py b/modeling_mixformer_sequential.py index 01d2c97..74c699e 100644 --- a/modeling_mixformer_sequential.py +++ b/modeling_mixformer_sequential.py @@ -756,9 +756,6 @@ class MixFormerSequentialForCausalLM(MixFormerSequentialPreTrainedModel): labels: Optional[torch.LongTensor] = None, **kwargs, ) -> CausalLMOutputWithPast: - if attention_mask is not None and self.training: - print("`attention_mask` is not supported during training. Using it might lead to unexpected results.") - if past_key_values is None and attention_mask is None: lm_logits = self.layers(input_ids) else: