delta

Author: sshleifer

src/transformers/models/falcon/modeling_falcon.py

@@ -23,6 +23,7 @@
 from torch import nn
 from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, LayerNorm, MSELoss
 from torch.nn import functional as F
+from transformers.activations import get_activation
 
 from ...modeling_attn_mask_utils import (
     AttentionMaskConverter,
@@ -733,7 +734,7 @@ def _upad_input(self, query_layer, key_layer, value_layer, attention_mask, query
             (max_seqlen_in_batch_q, max_seqlen_in_batch_k),
         )
 
-from transformers.activations import get_activation
+
 class FalconMLP(nn.Module):
     def __init__(self, config: FalconConfig):
         super().__init__()