[
https://issues.apache.org/jira/browse/MADLIB-1324?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16834128#comment-16834128
]
Frank McQuillan commented on MADLIB-1324:
-----------------------------------------
LGTM
{code}
SELECT madlib.madlib_keras_fit('mnist_train_packed', -- source_table
'model1', -- model
'model_arch_library', -- model_arch_table
1, -- model_arch_id
$$ loss='categorical_crossentropy',
optimizer='adam', metrics=['acc'] $$, -- compile_params
$$ batch_size=128, epochs=4 $$, -- fit_params
5, -- num_iterations
FALSE, -- use_gpu
NULL, -- validation_table
'Frank', -- name
'A test model' -- description
);
{code}
produces a model
> DL naming improvements for `dependent_varname` and `independent_varname`
> ------------------------------------------------------------------------
>
> Key: MADLIB-1324
> URL: https://issues.apache.org/jira/browse/MADLIB-1324
> Project: Apache MADlib
> Issue Type: Improvement
> Components: Module: Neural Networks
> Reporter: Frank McQuillan
> Priority: Minor
> Fix For: v1.16
>
>
> - use original column names from the source table (i.e., pre minibatch) in
> summary table, not the generic names `dependent_varname` and
> `independent_varname`
> - remove `dependent_varname` and `independent_varname` from the `fit()`
> interface since we know what these are from the minibatch summary table, and
> user must use minibatching for DL
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
