Update accelerate requirement from 0.27.2 to 0.28.0

[dependabot[bot]]

Updates the requirements on accelerate to permit the latest version.

Release notes

Sourced from accelerate's releases.

v0.28.0: DataLoaderConfig, XLA improvements, FSDP + QLORA foundations, Gradient Synchronization Tweaks, and Bug Fixes

Core

• Introduce a DataLoaderConfiguration and begin deprecation of arguments in the Accelerator

+from accelerate import DataLoaderConfiguration
+dl_config = DataLoaderConfiguration(split_batches=True, dispatch_batches=True)
-accelerator = Accelerator(split_batches=True, dispatch_batches=True)
+accelerator = Accelerator(dataloader_config=dl_config)

• Allow gradients to be synced each data batch while performing gradient accumulation, useful when training in FSDP by @​fabianlim in huggingface/accelerate#2531

from accelerate import GradientAccumulationPlugin
plugin = GradientAccumulationPlugin(
+    num_steps=2, 
    sync_each_batch=sync_each_batch
)
accelerator = Accelerator(gradient_accumulation_plugin=plugin)

Torch XLA

• Support for XLA on the GPU by @​anw90 in huggingface/accelerate#2176
• Enable gradient accumulation on TPU in huggingface/accelerate#2453

FSDP

• Support downstream FSDP + QLORA support through tweaks by allowing configuration of buffer precision by @​pacman100 in huggingface/accelerate#2544

launch changes

• Support mpirun for multi-cpu training by @​dmsuehir in huggingface/accelerate#2493

What's Changed

• Fix model metadata issue check by @​muellerzr in huggingface/accelerate#2435
• Use py 3.9 by @​muellerzr in huggingface/accelerate#2436
• Fix seedable sampler logic and expound docs by @​muellerzr in huggingface/accelerate#2434
• Fix tied_pointers_to_remove type by @​fxmarty in huggingface/accelerate#2439
• Make test assertions more idiomatic by @​akx in huggingface/accelerate#2420
• Prefer is_torch_tensor over hasattr for torch.compile. by @​PhilJd in huggingface/accelerate#2387
• Enable more Ruff lints & fix issues by @​akx in huggingface/accelerate#2419
• Fix warning when dispatching model by @​SunMarc in huggingface/accelerate#2442
• Make torch xla available on GPU by @​anw90 in huggingface/accelerate#2176
• Include pippy_file_path by @​muellerzr in huggingface/accelerate#2444
• [Big deprecation] Introduces a DataLoaderConfig by @​muellerzr in huggingface/accelerate#2441
• Check for None by @​muellerzr in huggingface/accelerate#2452
• Fix the pytest version to be less than 8.0.1 by @​BenjaminBossan in huggingface/accelerate#2461
• Fix wrong is_namedtuple implementation by @​fxmarty in huggingface/accelerate#2475
• Use grad-accum on TPU by @​muellerzr in huggingface/accelerate#2453
• Add pre-commit configuration by @​akx in huggingface/accelerate#2451
• Replace os.path.sep.join path manipulations with a helper by @​akx in huggingface/accelerate#2446
• DOC: Fixes to Accelerator docstring by @​BenjaminBossan in huggingface/accelerate#2443
• Context manager fixes by @​akx in huggingface/accelerate#2450

... (truncated)

Commits

• 9e72c61 Release: v0.28.0
• 354db5b Use uv instead of pip install for github CI (#2546)
• 92b1ad0 Update FSDP mixed precision setter to enable fsdp+qlora (#2544)
• 60bfdaa Allow Gradients to be Synced Each Data Batch While Performing Gradient Accumu...
• 16eb6d7 Remove extra double-dash in error message (#2541)
• c8acfa7 [docs] Troubleshoot (#2538)
• e70e3c8 Overdue email change... (#2534)
• bc8dfe3 init (#2438)
• e3d3242 Check if the buffers fit GPU memory after device map auto inferred (#2412)
• 10882ee Update link to dynamo/compile doc (#2533)
• Additional commits viewable in compare view

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

---

Dependabot commands and options

You can trigger Dependabot actions by commenting on this PR: - `@dependabot rebase` will rebase this PR - `@dependabot recreate` will recreate this PR, overwriting any edits that have been made to it - `@dependabot merge` will merge this PR after your CI passes on it - `@dependabot squash and merge` will squash and merge this PR after your CI passes on it - `@dependabot cancel merge` will cancel a previously requested merge and block automerging - `@dependabot reopen` will reopen this PR if it is closed - `@dependabot close` will close this PR and stop Dependabot recreating it. You can achieve the same result by closing it manually - `@dependabot show ignore conditions` will show all of the ignore conditions of the specified dependency - `@dependabot ignore this major version` will close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself) - `@dependabot ignore this minor version` will close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself) - `@dependabot ignore this dependency` will close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)