ccjoechou commented on a change in pull request #9328:
URL: https://github.com/apache/tvm/pull/9328#discussion_r732817799



##########
File path: python/tvm/relay/op/nn/_nn.py
##########
@@ -515,6 +515,35 @@ def 
compute_contrib_conv3d_winograd_weight_transform(attrs, inputs, out_dtype):
 reg.register_pattern("nn.max_pool2d", OpPattern.OUT_ELEMWISE_FUSABLE)
 
 
[email protected]_convert_op_layout("nn.max_pool2d")
+def convert_max_pool2d(attrs, inputs, tinfos, desired_layouts):
+    """Convert Layout pass registration for deformable conv2d op.

Review comment:
       Yes, for max pool 2d, global max pool 2d, avg pool 2d, and global avg 
pool 2d for now.




-- 
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.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


Reply via email to