However, release 1.1.1 was backwards compatible with release What is and is not the public API, see What is covered.įor example, release 1.0.0 introduced backwards incompatible changes from Non-experimental public API will continue to work unchanged. Code andĭata that worked with a previous minor release and which depends only on the MINOR: Backwards compatible features, speed improvements, etc. May be migratable to the newer release see However, in some cases existing TensorFlow graphs and checkpoints Worked with a previous major release will not necessarily work with the new MAJOR: Potentially backwards incompatible changes. Changes to each number have the following meaning: Each release version of TensorFlow has the form .įor example, TensorFlow version 1.2.3 has MAJOR version 1, MINOR version 2,Īnd PATCH version 3. TensorFlow follows Semantic Versioning 2.0 ( semver) for its To modify TensorFlow while preserving compatibility. Versions of TensorFlow (either for code or data), and for developers who want This document is for users who need backwards compatibility across different
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |