You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@madlib.apache.org by "Orhan Kislal (Jira)" <ji...@apache.org> on 2022/07/18 17:54:00 UTC

[jira] [Updated] (MADLIB-1478) DL: Passing in a custom function table to fit along with built in loss and metrics functions errors out

     [ https://issues.apache.org/jira/browse/MADLIB-1478?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Orhan Kislal updated MADLIB-1478:
---------------------------------
    Fix Version/s: v1.20.0
                       (was: v1.19.0)

> DL: Passing in a custom function table to fit along with built in loss and metrics functions errors out
> -------------------------------------------------------------------------------------------------------
>
>                 Key: MADLIB-1478
>                 URL: https://issues.apache.org/jira/browse/MADLIB-1478
>             Project: Apache MADlib
>          Issue Type: Bug
>          Components: Deep Learning
>            Reporter: Frank McQuillan
>            Priority: Minor
>             Fix For: v1.20.0
>
>
> Passing in a custom function table to fit along with built in loss and metrics functions errors out
> {code}
> SELECT madlib.load_top_k_accuracy_function('custom_function_table',
>                                            3);
> SELECT madlib.madlib_keras_fit('iris_data_packed', 'iris_model','iris_model_arch',1,
> $$loss='categorical_crossentropy',optimizer='Adam(lr=0.1)',metrics=['accuracy']$$,
> $$batch_size=10,epochs=1$$,1, FALSE, NULL, NULL, FALSE,NULL,NULL,'custom_function_table');
> DETAIL:  Traceback (most recent call last):\n  File "<string>", line 11, in __plpython_procedure_fit_transition_wide_359961\n  File "/Users/nikhilkak/workspace/madlib/build/src/ports/greenplum/6/modules/deep_learning/madlib_keras.py", line 579, in fit_transition_wide\n  File "/Users/nikhilkak/workspace/madlib/build/src/ports/greenplum/6/modules/deep_learning/madlib_keras.py", line 618, in fit_transition\n  File "/Users/nikhilkak/workspace/madlib/build/src/ports/greenplum/6/modules/deep_learning/madlib_keras.py", line 91, in get_init_model_and_sess\n  File "/Users/nikhilkak/workspace/madlib/build/src/ports/greenplum/6/modules/deep_learning/madlib_keras.py", line 551, in init_model\n  File "/Users/nikhilkak/workspace/madlib/build/src/ports/greenplum/6/modules/deep_learning/madlib_keras_wrapper.py", line 347, in compile_model\n  File "/Library/Python/2.7/site-packages/dill/_dill.py", line 283, in loads\n    return load(file, ignore, **kwds)\n  File "/Library/Python/2.7/site-packages/dill/_dill.py", line 278, in load\n    return Unpickler(file, ignore=ignore, **kwds).load()\n  File "/Library/Python/2.7/site-packages/dill/_dill.py", line 481, in load\n    obj = StockUnpickler.load(self)\n  File "/System/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/pickle.py", line 864, in load\n    dispatch[key](self)\nKeyError: \'{\'\n' (plpy_elog.c:121)  (seg0 slice1 127.0.0.1:6002 pid=85404)
> {code}



--
This message was sent by Atlassian Jira
(v8.20.10#820010)