zhiics commented on a change in pull request #4595: [Relay][Pass]Improve
memory_allocation pass to support multiple i/o dynamic kernels
URL: https://github.com/apache/incubator-tvm/pull/4595#discussion_r362108724
##########
File path: python/tvm/relay/memory_alloc.py
##########
@@ -220,7 +231,11 @@ def visit_call(self, call):
scope.let("shape_func", shape_call)
out_types = []
- out_types.append(call.checked_type)
+ if isinstance(ret_type, ty.TensorType):
Review comment:
Can we use `LinearizeRetType` here?
----------------------------------------------------------------
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]
With regards,
Apache Git Services
