drivanov commented on a change in pull request #16218: Improving performance of
argmax operator
URL: https://github.com/apache/incubator-mxnet/pull/16218#discussion_r328726787
##########
File path: tests/python/unittest/test_operator.py
##########
@@ -7604,6 +7612,161 @@ def _softmaxoutput_normalization(multi_output,
use_ignore, normalization):
multi_output, use_ignore, normalization)
+@with_seed()
+def test_argmax():
+ ci_test = True # Use False if you need to collect aggregated
information about argmax performance
Review comment:
Thanks for the idea! In fact, I myself did not like a lot of `if`'s in this
code. I have separated performance and unit test, but I think it would be
better to keep
```
def profiling_argmax():
```
in `tests/python/unittest/test_operator.py`
The truth is: I tried to add it to `tests/nightly/test_large_array.py`, but
it did not work for me. For some reason Python has some problem with this
modules first lines:
```
root@627d15c17420:/opt/mxnet# python
Python 3.5.2 (default, Nov 12 2018, 13:43:14)
[GCC 5.4.0 20160609] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import numpy as np
>>> import mxnet as mx
>>> from mxnet.test_utils import rand_ndarray, assert_almost_equal,
rand_coord_2d
>>> from mxnet import gluon, nd
>>> from tests.python.unittest.common import with_seed
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
ImportError: No module named 'tests.python'
>>>
```
----------------------------------------------------------------
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
