njayaram2 commented on a change in pull request #377: DL: Do not compile
params in predict and add gpu support
URL: https://github.com/apache/madlib/pull/377#discussion_r279136039
##########
File path: src/ports/postgres/modules/deep_learning/madlib_keras_predict.py_in
##########
@@ -112,6 +114,9 @@ def predict(schema_madlib, model_table, test_table, id_col,
class_values, intermediate_col, pred_col_name,
pred_col_type, is_response, MODULE_NAME)
+ segment_id = -1 if is_platform_pg() else '{0}.gp_segment_id'.format(
Review comment:
Realized that if we use the refactored function, we will have to call it
twice in fit function, once for train table and once for validation
table, since the function prefixes the table name for GPDB. I think it's
fine the way it was now, we can refactor it later if the use for this check
gets more ubiquitous. So reverted the refactoring commit, and will keep
it this way for now.
----------------------------------------------------------------
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
