Uses a TF2 SavedModel to construct a dense representation from text.

key A string or FeatureColumn identifying the input string data.
module_path A string path to the module. Can be a path to local filesystem or a handle.
output_key Name of the output item to return if the layer returns a dict. If the result is not a single value and an output_key is not specified, the feature column cannot infer the right output to use.
trainable Whether or not the Model is trainable. False by default, meaning the pre-trained weights are frozen. This is different from the ordinary tf.feature_column.embedding_column(), but that one is intended for training from scratch.

DenseColumn that converts from text input.