Version 6.0.2

[IvyZX]

What's Changed

• Refactor out dataclass transform that allows parent and name to be moved to the end of the argument list into a more general "kw_only_dataclasses" module. by @copybara-service in Refactor out dataclass transform that allows parent and name to be moved to the end of the argument list into a more general "kw_only_dataclasses" module. #2468
• Don't create reference cycles among Modules. by @levskaya in Don't create reference cycles among Modules. #2499
• adds perturb to docs by @cgarciae in adds perturb to docs #2511
• Add pre-commit hook to remove trailing white spaces by @cgarciae in Add pre-commit hook to remove trailing white spaces #2513
• Add extracting gradients section to capture intermediates guide by @cgarciae in Add extracting gradients section to capture intermediates guide #2515
• Make is_fully_replicated and is_fully_addressble a property rather than a method. by @copybara-service in Make is_fully_replicated and is_fully_addressble a property rather than a method. #2516
• Adds CallSetupUnboundModuleError by @cgarciae in Adds CallSetupUnboundModuleError #2496
• Adding documentation to Dropout around rng use by @zaxtax in Adding documentation to Dropout around rng use #2492
• no-op when double wrapping with struct.dataclass by @cgarciae in no-op when double wrapping with struct.dataclass #2505
• Add epub and pdf RDT formats by @cgarciae in Add epub and pdf RDT formats #2517
• Update landing page example by @cgarciae in Update landing page example #2366
• Use gfile.remove for files because it doesn't work on GCS files. by @IvyZX in Use gfile.remove for files because it doesn't work on GCS files. #2518
• Add save_checkpoint_multiprocess to api reference. by @IvyZX in Add save_checkpoint_multiprocess to api reference. #2519
• Add Tensorstore back to required dependencies. by @IvyZX in Add Tensorstore back to required dependencies. #2520
• Delete flax.optim.rst by @ppwwyyxx in Delete flax.optim.rst #2522
• Added an error for when we call init, init_with_output and apply on Module class. by @chiamp in Added an error for when we call init, init_with_output and apply on Module class. #2529
• Adding link to Bayesian Inference example that uses Flax by @zaxtax in Adding link to Bayesian Inference example that uses Flax #2521
• Added IncorrectPostInitOverrideError to capture incorrect post init overrides. by @copybara-service in Added IncorrectPostInitOverrideError to capture incorrect post init overrides. #2535
• Fix to use is_initializing for init-detection. by @yotarok in Fix to use is_initializing for init-detection. #2486
• add rng_collection argument to Dropout by @cgarciae in add rng_collection argument to Dropout #2540
• Cancel tests if other jobs fail by @cgarciae in Cancel tests if other jobs fail #2507
• Update Guides link in Flax README by @8bitmp3 in Update Guides link in Flax README #2544
• Pin jupytext version in requirements.txt by @IvyZX in Pin jupytext version in requirements.txt #2545
• Fix flax.linen.stochastic.Dropout by @dslisleedh in Fix flax.linen.stochastic.Dropout #2510
• Transfer Learning Guide by @cgarciae in Transfer Learning Guide #2394
• Update Flax Contributing.md by @8bitmp3 in Update Flax Contributing.md #2546
• cap dynamic scale to float32 max by @jheek in cap dynamic scale to float32 max #2553
• Remove optional import of jax.experimental.array for older JAX versions. by @copybara-service in Remove optional import of jax.experimental.array for older JAX versions. #2552
• Update examples.rst with denoising-diffusion-flax by @yiyixuxu in Update examples.rst with denoising-diffusion-flax #2487
• return None if no _parent_ref is set by @cgarciae in return None if no _parent_ref is set #2548
• Add a documentation page on checkpointing by @IvyZX in Add a documentation page on checkpointing #2530
• Lint Flax Contributing guide by @8bitmp3 in Lint Flax Contributing guide #2560
• Remove extra metadata in Checkpointing guide by @8bitmp3 in Remove extra metadata in Checkpointing guide #2559
• Update getting_started.ipynb and getting_started.md by @chiamp in Update getting_started.ipynb and getting_started.md #2563
• Remove unused svn dependency by @8bitmp3 in Remove unused svn dependency  #2574
• Fix pytype check in checkpoints.py by @IvyZX in Fix pytype check in checkpoints.py #2592
• Add new 🔪 Flax - The Sharp Bits 🔪 Dropout and randomness by @8bitmp3 in Add new 🔪 Flax - The Sharp Bits 🔪 Dropout and randomness #2593
• Fixes launch_gce.sh with imagenet example. by @andsteing in Fixes launch_gce.sh with imagenet example. #2598
• Added test to check for Variable warning. by @chiamp in Added test to check for Variable warning. #2610
• Release version 0.6.2 by @IvyZX in Release version 0.6.2 #2613

New Contributors

• @zaxtax made their first contribution in Adding documentation to Dropout around rng use #2492
• @ppwwyyxx made their first contribution in Delete flax.optim.rst #2522
• @chiamp made their first contribution in Added an error for when we call init, init_with_output and apply on Module class. #2529
• @yotarok made their first contribution in Fix to use is_initializing for init-detection. #2486
• @yiyixuxu made their first contribution in Update examples.rst with denoising-diffusion-flax #2487

Full Changelog: v0.6.1...v0.6.2

---

This discussion was created from the release Version 6.0.2.