Coverage for /pythoncovmergedfiles/medio/medio/usr/local/lib/python3.8/site-packages/tensorflow/_api/v2/compat/v1/saved_model/loader/__init__.py: 100%
4 statements
« prev ^ index » next coverage.py v7.4.0, created at 2024-01-03 07:57 +0000
« prev ^ index » next coverage.py v7.4.0, created at 2024-01-03 07:57 +0000
1# This file is MACHINE GENERATED! Do not edit.
2# Generated by: tensorflow/python/tools/api/generator/create_python_api.py script.
3"""Loader functionality for SavedModel with hermetic, language-neutral exports.
5Load and restore capability for a SavedModel, which may include multiple meta
6graph defs. Each SavedModel is associated with a single checkpoint. Each meta
7graph def is saved with one or more tags, which are used to identify the exact
8meta graph def to load.
10The `load` operation requires the session in which to restore the graph
11definition and variables, the tags used to identify the meta graph def to
12load and the location of the SavedModel.
14Upon a load, the subset of variables and assets supplied as part of the specific
15meta graph def, will be restored into the supplied session. The values of the
16variables though will correspond to the saved values from the first meta graph
17added to the SavedModel using `add_meta_graph_and_variables(...)` in
18`builder.py`.
20Typical usage:
22```python
23...
24builder = tf.compat.v1.saved_model.builder.SavedModelBuilder(export_dir)
26with tf.compat.v1.Session(graph=tf.Graph()) as sess:
27 ...
28 builder.add_meta_graph_and_variables(sess,
29 ["foo-tag"],
30 signature_def_map=foo_signatures,
31 assets_collection=foo_assets)
32...
34with tf.compat.v1.Session(graph=tf.Graph()) as sess:
35 ...
36 builder.add_meta_graph(["bar-tag", "baz-tag"],
37 assets_collection=bar_baz_assets)
38...
40builder.save()
42...
43with tf.compat.v1.Session(graph=tf.Graph()) as sess:
44 tf.compat.v1.saved_model.loader.load(sess, ["foo-tag"], export_dir)
45 ...
47```
49"""
51import sys as _sys
53from tensorflow.python.saved_model.loader_impl import load
54from tensorflow.python.saved_model.loader_impl import maybe_saved_model_directory