spaCy - Util.load_model_from_path

It is introduced in version 2.0. As the name suggests, this function will load a model from a data directory path.

It works by creating the language class and pipeline based on directory’s meta.json. Later on, it will call from_disk with the specified path. With this utility function, we can also easily test a new model that is yet to be packaged.

Arguments

The table below explains its arguments −

NAME	TYPE	DESCRIPTION
model_path	unicode	It is the path to the model the data directory.
disable	List	It represents the names of pipeline components to disable.
meta	Dict	It is model meta data. If this value is false, spaCy will try to load the meta from a meta.json in the same directory.

Example

An example for util.load_model_from_path() function is given below −

import spacy
nlp_model = spacy.util.load_model_from_path ("/path/to/data")

spacy_util_get_data_path.htm