Icon

01_​Text_​Generation_​Fairy_​Tales_​Training

Generate Text Using a Many-To-One LSTM Network (Training)

The workflow builds, trains, and saves an RNN with an LSTM layer to generate new fictive fairy tales. The brown nodes define the network structure. The "Pre-Processing and Encoding" part of the workflow reads the fairy tales, index-encodes them, and creates semi-overlapping sequences. The Keras Network Learner node trains the network using the index-encoded fairy tales. Finally, the trained network is converted into a TensorFlow model, and saved to a file.

Pre Processing and Encoding Train and Save Network Define Network Structure Apply dictionaryCreate overlappingsequencesfor regularization?, dictionary sizeCreate inputcollectionCreate targetcollectionDelete superfluouscolumns one column,each char in single rowactivation softmaxunits = dictionary size Cell Replacer Lag Column Keras Dropout Layer Keras Input Layer Keras NetworkLearner Create CollectionColumn Create CollectionColumn Keras to TensorFlowNetwork Converter Resort Columns Column Filter Reshape Text Keras NetworkWriter TensorFlowNetwork Writer Keras LSTM Layer Keras Dense Layer Read and ExtractFairy Tales Create and SaveDictionary Pre Processing and Encoding Train and Save Network Define Network Structure Apply dictionaryCreate overlappingsequencesfor regularization?, dictionary sizeCreate inputcollectionCreate targetcollectionDelete superfluouscolumns one column,each char in single rowactivation softmaxunits = dictionary size Cell Replacer Lag Column Keras Dropout Layer Keras Input Layer Keras NetworkLearner Create CollectionColumn Create CollectionColumn Keras to TensorFlowNetwork Converter Resort Columns Column Filter Reshape Text Keras NetworkWriter TensorFlowNetwork Writer Keras LSTM Layer Keras Dense Layer Read and ExtractFairy Tales Create and SaveDictionary

Nodes

Extensions

Links