[Encoder Decoder] Add flash_attn kernel support for encoder-decoder models (#9559)
This commit is contained in:
@@ -624,8 +624,6 @@ class BartEncoder(nn.Module):
|
||||
Decoder output torch.Tensor
|
||||
"""
|
||||
# retrieve input_ids and inputs_embeds
|
||||
|
||||
input_ids = input_ids.view(-1, input_ids.shape[-1])
|
||||
inputs_embeds = self.embed_tokens(input_ids)
|
||||
|
||||
embed_pos = self.embed_positions(
|
||||
|
||||
Reference in New Issue
Block a user