Neutron3529 commented on a change in pull request #19748:
URL: https://github.com/apache/incubator-mxnet/pull/19748#discussion_r558865892
##########
File path: tests/python/unittest/test_gluon_data.py
##########
@@ -158,10 +158,11 @@ def __getitem__(self, key):
def test_multi_worker():
data = Dataset()
for thread_pool in [True, False]:
- loader = gluon.data.DataLoader(data, batch_size=1, num_workers=5,
thread_pool=thread_pool)
- for i, batch in enumerate(loader):
- assert (batch.asnumpy() == i).all()
-
+ for auto_reload in [True, False]:
Review comment:
why and how...
Actually I just do not know how to use `pytest.mark.parametrize`.
and, I saw such code at line 263:
```
for epoch in range(1):
for i, data in enumerate(loader):
```
it seems that a `for-loop` is acceptable even if there is only one parameter
in the iterator.
----------------------------------------------------------------
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]
