leandron commented on a change in pull request #6278:
URL: https://github.com/apache/incubator-tvm/pull/6278#discussion_r470486489
##########
File path: python/tvm/relay/frontend/keras.py
##########
@@ -123,10 +124,11 @@ def _convert_advanced_activation(inexpr, keras_layer,
etab):
if isinstance(axis, list):
raise tvm.error.OpAttributeUnImplemented(
'Softmax with axes {} is not supported.'.format(axis))
- if axis == -1:
- axis = 1
- else:
- axis = axis + 1 if axis < dims - 1 else 1
+ if etab.data_layout == 'NCHW':
Review comment:
Maybe silly question but need to ask: are there no expected adjusts to
be made on `axis` here when layout is not `NCHW`?
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
