freshcrate
Skin:/
Home > Frameworks > orbax-checkpoint

orbax-checkpoint

Orbax Checkpoint

Why this rank:Strong adoptionRecent releaseHealthy release cadence

Description

# Orbax Checkpointing `pip install orbax-checkpoint` (latest PyPi release) OR `pip install 'git+https://github.com/google/orbax/#subdirectory=checkpoint'` (from this repository, at HEAD) `import orbax.checkpoint` Orbax includes a checkpointing library oriented towards JAX users, supporting a variety of different features required by different frameworks, including asynchronous checkpointing, various types, and various storage formats. We aim to provide a highly customizable and composable API which maximizes flexibility for diverse use cases.

Release History

VersionChangesUrgencyDate
v0.12.0### Added - Add `LatestDuration` preservation policy to support duration-based (TTL) checkpoint retention. - Add `EveryNStepsClosest` preservation policy, which preserves checkpoints at steps closest to multiples of `interval_steps`. - Add `SerializationStatusCallback` for save operations to track transfer and disk write events. While primarily intended as lightweight progress handles, this callback system allows the potential to inject custom logic for advanced monitoring and custom actions High6/2/2026
v0.11.40### Removed - #v1 Remove numpy and pytorch layouts, as these are insecure and not worth the effort in maintaining. Safetensors is instead the recommended conversion case. - #v1 Remove `LeafHandler` as a user-exposed layer (it remains as an internal layer). ### Added - #v1 Allow a context to be default-configured for all `Checkpointer` operations. ### Changed - Upgrade to TensorStore>=0.1.84 (and Python>=3.11 as a consequence.)High5/19/2026
v0.11.39### Added - Tensorstore non-atomic file I/O locking on local/POSIX-compatible filesystems to avoid unnecessary renames.High5/6/2026
v0.11.37### Added - Add colocated runtime helpers for Pathways MTC. - #v1 Centralize `StorageOptions` into `ArrayOptions` and implement field-level merging. - Add Patch for Pathways CPU ids. - Add sidecar MTC backend support - Add Colocated Python Sidecar MTC Checkpoint Manager - Add Colocated Python Orchestration handling - #v1 Add `LeafHandler` as a `CheckpointableHandler`, so that ordinary PyTree leaves can also be saved as individual checkpointables. - Move MTC files to multi_tier_checkpHigh5/5/2026
0.11.36Imported from PyPI (0.11.36)Low4/21/2026
v0.11.36### Added - #v1 Add auto loading functionality to `load_pytree` and `pytree_metadata`.Medium4/14/2026
v0.11.36### Added - #v1 Add auto loading functionality to `load_pytree` and `pytree_metadata`.Medium4/14/2026
v0.11.36### Added - #v1 Add auto loading functionality to `load_pytree` and `pytree_metadata`.Medium4/14/2026
v0.11.36### Added - #v1 Add auto loading functionality to `load_pytree` and `pytree_metadata`.Medium4/14/2026
v0.11.36### Added - #v1 Add auto loading functionality to `load_pytree` and `pytree_metadata`.Medium4/14/2026
v0.11.36### Added - #v1 Add auto loading functionality to `load_pytree` and `pytree_metadata`.Medium4/14/2026
v0.11.36### Added - #v1 Add auto loading functionality to `load_pytree` and `pytree_metadata`.Medium4/14/2026
v0.11.36### Added - #v1 Add auto loading functionality to `load_pytree` and `pytree_metadata`.Medium4/14/2026
v0.11.36### Added - #v1 Add auto loading functionality to `load_pytree` and `pytree_metadata`.Medium4/14/2026
v0.11.36### Added - #v1 Add auto loading functionality to `load_pytree` and `pytree_metadata`.Medium4/14/2026
v0.11.36### Added - #v1 Add auto loading functionality to `load_pytree` and `pytree_metadata`.Medium4/14/2026
v0.11.36### Added - #v1 Add auto loading functionality to `load_pytree` and `pytree_metadata`.Medium4/14/2026
v0.11.36### Added - #v1 Add auto loading functionality to `load_pytree` and `pytree_metadata`.Medium4/14/2026
v0.11.35### Added - Add PyTorch DCP (Distributed Checkpoint) to the benchmark suite, with GCS connector and fsspec options. - Eliminate unadvertised flax dependency from tree/utils. ### Fixed - #v1 Fix import path for `step` in `orbax.checkpoint.experimental.v1.path`.Medium4/13/2026
v0.11.34Release v0.11.34Medium4/8/2026
v0.11.33Release v0.11.33Low2/18/2026
v0.11.32Release v0.11.32Low1/21/2026
v0.11.30Release v0.11.30Low11/27/2025
v0.11.28Release v0.11.28Low11/6/2025
v0.11.27Release v0.11.27Low11/5/2025
v0.11.26Release v0.11.26Low10/22/2025
v0.11.25Release v0.11.25Low9/11/2025
v0.11.24Release v0.11.24Low8/28/2025
v0.11.23Release v0.11.23Low8/18/2025
v0.11.22Release v0.11.22Low8/14/2025
v0.11.21Release v0.11.21Low8/7/2025
v0.11.20Release v0.11.20Low7/28/2025
v0.11.19Release v0.11.19Low7/9/2025
v0.11.18Release v0.11.18Low7/1/2025
v0.11.17Release v0.11.17Low6/30/2025
v0.11.16Release v0.11.16Low6/18/2025
v0.11.15Release v0.11.15Low6/13/2025
v0.11.14Release v0.11.14Low6/10/2025
v0.11.13Release v0.11.13Low5/1/2025
v0.11.12Release v0.11.12Low4/9/2025
v0.11.11Release v0.11.11Low4/7/2025
v0.11.10Release v0.11.10Low3/20/2025
v0.11.9Release v0.11.9Low3/17/2025
v0.11.8Release v0.11.8Low3/7/2025
v0.11.7Release v0.11.7Low3/7/2025
v0.11.6Release v0.11.6Low2/21/2025
v0.11.5Release v0.11.5Low2/11/2025
v0.11.4Release v0.11.4Low2/7/2025
v0.11.3Release v0.11.3Low2/6/2025
v0.11.2Release v0.11.2Low2/4/2025
v0.11.1Release v0.11.1Low1/28/2025
v0.11.0Release v0.11.0Low12/30/2024
v0.10.3Release v0.10.3Low12/19/2024
v0.10.2Release v0.10.2Low12/4/2024
v0.10.1Release v0.10.1Low11/22/2024
v0.10.0Release v0.10.0Low11/22/2024
v0.9.1Release v0.9.1Low11/11/2024
v0.9.0Release v0.9.0Low11/8/2024
v0.8.0Release v0.8.0Low10/29/2024
v0.7.0Release v0.7.0Low10/7/2024
v0.6.4Release v0.6.4Low9/17/2024
v0.6.3Release v0.6.3Low9/10/2024
v0.1.7### Added - Support for OCDBT driver in Tensorstore.Low3/29/2023
v0.1.6### Fixed - Small bug fixes.Low3/22/2023
v0.1.5### Added - Use a more precise timestamp when generating temporary directory names to permit more than one concurrent checkpointing attempt per second. - Automatic import of nest_asyncio.Low3/17/2023
v0.1.4### Added - Support for generic transformation function in PyTreeCheckpointHandler. - Support n-digit checkpoint step format. ### Fixed - Eliminate Flax dependency to fix circular dependency problem.Low3/15/2023
v0.1.3### Added - `sharding` option on `ArrayRestoreArgsLow3/3/2023

Dependencies & License Audit

Loading dependencies...

Similar Packages

tensorflowTensorFlow is an open source machine learning framework for everyone.2.21.0
ctranslate2Fast inference engine for Transformer modelsv4.7.2
xgrammarEfficient, Flexible and Portable Structured Generationv0.2.1
apache-tvm-ffitvm ffiv0.1.11
magikaA tool to determine the content type of a file with deep learningcli/v1.1.0

More from pypi

markitdownUtility tool for converting various files to Markdown
fastapiFastAPI framework, high performance, easy to learn, fast to code, ready for production
djangoA high-level Python web framework that encourages rapid development and clean, pragmatic design.
flaskA simple framework for building complex web applications.

More in Frameworks

langchainThe agent engineering platform
deer-flowAn open-source long-horizon SuperAgent harness that researches, codes, and creates. With the help of sandboxes, memories, tools, skill, subagents and message gateway, it handles different levels of ta
tqdmFast, Extensible Progress Meter
simBuild, deploy, and orchestrate AI agents. Sim is the central intelligence layer for your AI workforce.