mmlspark.cntk package¶

Submodules¶

mmlspark.cntk.CNTKModel module¶

class mmlspark.cntk.CNTKModel.CNTKModel(java_obj=None, batchInput=True, convertOutputToDenseVector=True, feedDict={'ARGUMENT_0': 'ARGUMENT_0'}, fetchDict={'OUTPUT_0': 'OUTPUT_0'}, miniBatcher=None, model=None, shapeOutput=False)[source]¶

Bases: mmlspark.cntk._CNTKModel._CNTKModel

Parameters

SparkSession (SparkSession) – The SparkSession that will be used to find the model
location (str) – The location of the model, either on local or HDFS

getInputCol()[source]¶

getInputNode()[source]¶

getInputNodeIndex()[source]¶

getInputShapes()[source]¶

getOutputCol()[source]¶

getOutputNode()[source]¶

getOutputNodeIndex()[source]¶

rebroadcastCNTKModel(sparkSession)[source]¶

setFeedDict(dict)[source]¶

Parameters: feedDict – Map of CNTK Variable names (keys) and Column Names (values)

setFetchDict(dict)[source]¶

Parameters: fetchDict – Map of Column Names (keys) and CNTK Variable names (values)

setInputCol(n)[source]¶

setInputNode(n)[source]¶

setInputNodeIndex(n)[source]¶

setMiniBatchSize(n)[source]¶

setModelLocation(location)[source]¶

setOutputCol(n)[source]¶

setOutputNode(n)[source]¶

setOutputNodeIndex(n)[source]¶

mmlspark.cntk.ImageFeaturizer module¶

class mmlspark.cntk.ImageFeaturizer.ImageFeaturizer(java_obj=None, cntkModel=None, cutOutputLayers=1, dropNa=True, inputCol=None, layerNames=None, outputCol='ImageFeaturizer_757e969e1f68_output')[source]¶

Bases: mmlspark.cntk._ImageFeaturizer._ImageFeaturizer

Parameters

SparkSession (SparkSession) – The SparkSession that will be used to find the model
ocation (str) – The location of the model, either on local or HDFS

setMiniBatchSize(size)[source]¶

setModel(modelSchema)[source]¶

setModelLocation(location)[source]¶

Module contents¶

MMLSpark is an ecosystem of tools aimed towards expanding the distributed computing framework Apache Spark in several new directions. MMLSpark adds many deep learning and data science tools to the Spark ecosystem, including seamless integration of Spark Machine Learning pipelines with Microsoft Cognitive Toolkit (CNTK), LightGBM and OpenCV. These tools enable powerful and highly-scalable predictive and analytical models for a variety of datasources.

MMLSpark also brings new networking capabilities to the Spark Ecosystem. With the HTTP on Spark project, users can embed any web service into their SparkML models. In this vein, MMLSpark provides easy to use SparkML transformers for a wide variety of Microsoft Cognitive Services. For production grade deployment, the Spark Serving project enables high throughput, sub-millisecond latency web services, backed by your Spark cluster.

MMLSpark requires Scala 2.11, Spark 2.4+, and Python 3.5+.