[ 
https://issues.apache.org/jira/browse/MADLIB-1371?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16906707#comment-16906707
 ] 

Frank McQuillan commented on MADLIB-1371:
-----------------------------------------

madlib=# SELECT madlib.madlib_keras_fit('iris_train_packed',   -- source table
madlib(#                                'iris_model',          -- model output 
table
madlib(#                                'model_arch_library',  -- model arch 
table
madlib(#                                 NULL,                    -- model arch 
id
madlib(#                                 $$ loss='categorical_crossentropy', 
optimizer='adam', metrics=['accuracy'] $$,  -- compile_params
madlib(#                                 $$ batch_size=5, epochs=3 $$,  -- 
fit_params
madlib(#                                 10                    -- num_iterations
madlib(#                               );
ERROR:  plpy.Error: madlib_keras_fit: Invalid model architecture ID. 
(plpython.c:5038)
CONTEXT:  Traceback (most recent call last):
  PL/Python function "madlib_keras_fit", line 21, in <module>
    madlib_keras.fit(**globals())
  PL/Python function "madlib_keras_fit", line 42, in wrapper
  PL/Python function "madlib_keras_fit", line 65, in fit
  PL/Python function "madlib_keras_fit", line 232, in __init__
  PL/Python function "madlib_keras_fit", line 256, in _validate_input_args
  PL/Python function "madlib_keras_fit", line 134, in validate_model_arch_table
  PL/Python function "madlib_keras_fit", line 96, in _assert
PL/Python function "madlib_keras_fit"


> DL: Exception during madlib_keras_fit when model_arch_id is passed as NULL
> --------------------------------------------------------------------------
>
>                 Key: MADLIB-1371
>                 URL: https://issues.apache.org/jira/browse/MADLIB-1371
>             Project: Apache MADlib
>          Issue Type: Bug
>          Components: Deep Learning
>            Reporter: Nikhil
>            Priority: Major
>             Fix For: v1.17
>
>
> While calling madlib_keras_fit, if we pass in the model arch id as NULL, we 
> get the following exception
> {code}
> SELECT madlib_keras_fit(
>  'cifar_10_sample_batched',
>  'keras_saved_out',
>  'model_arch',
>  NULL, --null model arch
>  $$ optimizer=SGD(lr=0.01, decay=1e-6, nesterov=True), 
> loss='categorical_crossentropy', metrics=['mae']$$::text,
>  $$ batch_size=2, epochs=1, verbose=0 $$::text,
>  3,
>  NULL);
> LINE 1: ...el_arch, model_weights FROM model_arch WHERE model_id = None
>                                                                    ^
> QUERY:  SELECT model_arch, model_weights FROM model_arch WHERE model_id = None
> CONTEXT:  Traceback (most recent call last):
>   PL/Python function "madlib_keras_fit", line 21, in <module>
>     madlib_keras.fit(**globals())
>   PL/Python function "madlib_keras_fit", line 42, in wrapper
>   PL/Python function "madlib_keras_fit", line 95, in fit
> PL/Python function "madlib_keras_fit"
> {code}
>  
>  We should instead error out gracefully and print a more meaningful error 
> message 
>  
>  



--
This message was sent by Atlassian JIRA
(v7.6.14#76016)

Reply via email to