reductionista commented on a change in pull request #524:
URL: https://github.com/apache/madlib/pull/524#discussion_r589817942
##########
File path:
src/ports/postgres/modules/deep_learning/test/unit_tests/test_madlib_keras.py_in
##########
@@ -1259,6 +1259,26 @@ class MadlibKerasWrapperTestCase(unittest.TestCase):
self.subject.parse_and_validate_compile_params(test_str)
self.assertIn('invalid optimizer', str(error.exception))
+ def test_parse_callbacks_pass(self):
+ test_str = """'[TensorBoard(log_dir="/tmp/logs/fit")]'"""
Review comment:
Looks like you asked this in another comment already. I agree with
@orhankislal 's response:
"I don't think we have to test for every keras parameter combination. The
log_dir is tested because it is error-prone with arbitrary paths but a numeric
value like batch_size either works or keras throws an appropriate error."
----------------------------------------------------------------
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:
us...@infra.apache.org
