Monkeybread Software - MBS FileMaker Plugin: CoreML.Description

Topics All MacOS (Only) Windows (Only) Linux (Only, Not) iOS (Only, Not)
Components Crossplatform Mac & Win Server Client Old Deprecated Guides Examples Videos
New in version: 9.3 9.4 9.5 10.0 10.1 10.2 10.3 10.4 10.5 10.6 Statistic FMM Blog

CoreML.Description

Queries the description of the model.

Component	Version	macOS	Windows	Linux	Server	FileMaker iOS SDK
CoreML	7.4	Yes	No	No	Yes, macOS only	Yes

MBS( "CoreML.Description"; CoreMLRef ) More

Parameters

Parameter Description Example

CoreMLRef The reference number to the model. $Model

Parameter	Description	Example
CoreMLRef	The reference number to the model.	$Model

Result

Returns JSON or error.

Description

Queries the description of the model.
This is returned as a JSON block.
Common keys include predictedFeatureName, predictedProbabilitiesName, metadata, inputDescriptionsByName and outputDescriptionsByName.

Added isUpdatable key in v10.0 to indicate whether model can be updated.
Includes trainingInputDescriptionsByName and parameterDescriptionsByKey For MacOS 10.15 to provide details on training input and parameters.

Examples

Query description:

Set Field [ CoreML::Model Description ; MBS("CoreML.Description"; $$ML) ]

Example result:
{
"predictedProbabilitiesName" : "classLabelProbs",
"predictedFeatureName" : "classLabel",
"inputDescriptionsByName" : {
"image" : {
"optional" : false,
"type" : "Image",
"imageConstraint" : {
"pixelsWide" : 224,
"pixelFormatTypeName" : "32BGRA",
"pixelFormatType" : 1111970369,
"pixelFormatTypeDescription" : "32 bit BGRA",
"pixelsHigh" : 224
},
"name" : "image"
}
},
"outputDescriptionsByName" : {
"classLabel" : {
"optional" : false,
"type" : "String",
"name" : "classLabel"
},
"classLabelProbs" : {
"optional" : false,
"dictionaryConstraint" : {
"keyType" : "String"
},
"type" : "Dictionary",
"name" : "classLabelProbs"
}
},
"metadata" : {
"MLModelVersionStringKey" : "",
"MLModelDescriptionKey" : "Detects the dominant objects present in an image from a set of 1000 categories such as trees, animals, food, vehicles, person etc. The top-5 error from the original publication is 7.8%.",
"MLModelCreatorDefinedKey" : {

},
"MLModelAuthorKey" : "Original Paper: Kaiming He and Xiangyu Zhang and Shaoqing Ren and Jian Sun. Keras Implementation: François Chollet",
"MLModelLicenseKey" : "MIT License. More information available at https:\/\/github.com\/fchollet\/keras\/blob\/master\/LICENSE"
}
}

Load model to get description:

Set Variable [ $model ; Value: MBS( "CoreML.OpenModel"; "/Users/cs/Desktop/UpdatableDrawingClassifier.mlmodelc" ) ]
Set Field [ test::ModelDescription ; MBS( "CoreML.Description"; $model ) ]
Set Variable [ $r ; Value: MBS( "CoreML.Release"; $model ) ]

Example Databases

Blog Entries

Release notes

Version 10.0
- Improved CoreML.Description to include newer attributes for MacOS 10.15.

Created 16th September 2017, last changed 16th December 2019

CoreML.CompileModel - CoreML.GetUsesCPUOnly

Feedback: Report problem or ask question.