yzhliu commented on a change in pull request #6428:
URL: https://github.com/apache/incubator-tvm/pull/6428#discussion_r485338529
##########
File path: python/tvm/topi/nn/conv2d_transpose.py
##########
@@ -98,9 +98,9 @@ def declaration_conv2d_transpose_impl(data, kernel, strides,
padding, out_dtype,
out_h = simplify(in_h - filter_h + 1 + output_padding[0])
out_w = simplify(in_w - filter_w + 1 + output_padding[1])
- dc = tvm.reduce_axis((0, in_c), name='dc')
- dh = tvm.reduce_axis((0, filter_h), name='dh')
- dw = tvm.reduce_axis((0, filter_w), name='dw')
+ dc = tvm.te.reduce_axis((0, in_c), name='dc')
+ dh = tvm.te.reduce_axis((0, filter_h), name='dh')
+ dw = tvm.te.reduce_axis((0, filter_w), name='dw')
Review comment:
seems we can omit "tvm" and do `te.reduce_axis`. can we add test case to
prevent from regression?
----------------------------------------------------------------
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]
