jwfromm commented on a change in pull request #6616:
URL: https://github.com/apache/incubator-tvm/pull/6616#discussion_r499005260
##########
File path: tests/python/topi/python/test_topi_batch_matmul.py
##########
@@ -67,10 +67,13 @@ def check_device(device, ctx):
@tvm.testing.uses_gpu
def test_batch_matmul():
- verify_batch_matmul(1, 16, 16, 32)
- verify_batch_matmul(5, 16, 16, 32)
- verify_batch_matmul(5, 16, 20, 32)
- verify_batch_matmul(30, 16, 20, 32)
+ verify_batch_matmul(1, 1, 16, 16, 32)
+ verify_batch_matmul(5, 5, 16, 16, 32)
+ verify_batch_matmul(5, 5, 16, 20, 32)
+ verify_batch_matmul(30, 30, 16, 20, 32)
+ # Test batch broadcasting.
+ verify_batch_matmul(1, 5, 16, 16, 32)
+ verify_batch_matmul(5, 1, 16, 16, 32)
Review comment:
Its unusual to add a test that intentionally fails to the topi testing
suite, it doesnt look any there are any other examples of that. We could
instead add it to the relay testing of batch_matmul if youd like.
----------------------------------------------------------------
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]
