Domino Valdano created MADLIB-1332:
--------------------------------------
Summary: DL: Support mini-batched validation data for fit/evaluate
Key: MADLIB-1332
URL: https://issues.apache.org/jira/browse/MADLIB-1332
Project: Apache MADlib
Issue Type: Improvement
Components: Deep Learning
Reporter: Domino Valdano
Fix For: v1.16
Currently, `keras_evaluate()` is implemented by calling
`internal_keras_evaluate() as a UDF. This requires the validation table passed
to `keras_fit()` to be in a format with only 1 image per row, even though the
training table is in a different format, with a batch of images in every row.
This is potentially confusing and cumbersome for users to deal with, and based
on some preliminary testing it seems that passing only 1 image at a time to
`keras_evaluate()` is also slowing down performance.
We can solve this by converting `internal_keras_evaluate()` into a UDA, so that
it runs on a minibatched validation table in the same form as the training
table.
Tasks:
* Convert the {{internal_keras_evaluate}} UDF to a UDA and perform weighted
averaging of loss and accuracy.
* Since x and y will now be minibatched, we don't need to add another
dimension to {{x and y}} np arrays in {{internal_keras_evaluate}}.
* Compare UDF to UDA and verify that the UDA results in a speed improvement
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
