[PyTorch] forward attention_type in MultiHeadAttention (#621)

[PyTorch] fix forward attention_type in MultiheadAttention
Signed-off-by: Markus Schnoes <markus.schnoes@gmx.de>
Co-authored-by: Kirthi Shankar Sivamani <ksivamani@nvidia.com>

transformer_engine/pytorch/attention.py

@@ -3090,6 +3090,7 @@ class MultiheadAttention(torch.nn.Module):
             sequence_parallel=sequence_parallel,
             tp_group=tp_group,
             layer_number=self.layer_number,
+            attention_type=self.attention_type,
         )
 
         # Linear