diff --git a/transformers/modeling_bert.py b/transformers/modeling_bert.py index f982364f5e6..a5b21510aaf 100644 --- a/transformers/modeling_bert.py +++ b/transformers/modeling_bert.py @@ -322,7 +322,7 @@ class BertLayer(nn.Module): def __init__(self, config): super(BertLayer, self).__init__() self.self_attention = BertAttention(config) - if config.get('is_decoder', False): + if config.get("is_decoder", False): self.attention = BertAttention(config) self.intermediate = BertIntermediate(config) self.output = BertOutput(config)