[GitHub] [madlib] khannaekta commented on a change in pull request #516: DL: Implement caching for fit_multiple_model

[GitBox <gi...@apache.org>]

khannaekta commented on a change in pull request #516:
URL: https://github.com/apache/madlib/pull/516#discussion_r492339498



##########
File path: src/ports/postgres/modules/deep_learning/test/madlib_keras_model_selection.sql_in
##########
@@ -580,6 +741,44 @@ FROM (SELECT count(*) cnt FROM iris_multiple_model_info
 WHERE compile_params = $MAD$loss='categorical_crossentropy', optimizer='Adam(lr=0.01)', metrics=['accuracy']$MAD$::text
 AND fit_params = $MAD$batch_size=32, epochs=1$MAD$::text) info;
 
+-- Test with caching when number of configs(4) larger than number of segments(3)
+DROP TABLE if exists iris_multiple_model, iris_multiple_model_summary, iris_multiple_model_info;
+SELECT madlib_keras_fit_multiple_model(
+	'iris_data_packed',
+	'iris_multiple_model',
+	'mst_table_4row',
+	3,
+	FALSE, NULL, NULL, NULL, NULL, NULL,
+	TRUE
+);

Review comment:
       Yes, this will still pass as it asserts on the number of models trained. Eventhough  for a cluster size > 3 segments, it might not test the use case intended in this specific test but it would not fail.




----------------------------------------------------------------
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:
users@infra.apache.org