canonical

Documentation

Overview

Package canonical is an alias for the `model/io` package that also links all the canonical model support along. Most of the time one wants to use this package instead of `model/io`. But to decrease code bloat, one can also depend on `model/io` and only the specific type of model desired.

Index

• Variables

Variables

var (
	// LoadModel loads a model from disk.
	// This is just an alias, see implementation in `model/io.go`
	LoadModel = model_io.LoadModel

	// LoadModelWithPrefix loads a model with a prefix from disk.
	// This is just an alias, see implementation in `model/io.go`
	//
	// The "prefix" is a string append to the name of all the files in the model. Using a prefix make
	// it possible to store multiple models in the same directory without sub-directories. Models
	// created by TensorFlow Decision Forests generally have prefix.
	LoadModelWithPrefix = model_io.LoadModelWithPrefix

	// DetectFilePrefix detect the prefix of the model.
	// This function is similar as "DetectFilePrefix" in `model_library.cc`.
	// This is just an alias, see implementation in `model/io.go`
	DetectFilePrefix = model_io.DetectFilePrefix
)