Pull Requests huggingface/accelerate

#4039 opened 2026-05-12 08:34 by 1fanwang

Improve distributed training example usability for automatic gradient accumulation

#4038 opened 2026-05-11 17:55 by Pritiks23

docs: Fix docstring of `fsdp2_prepare_auto_wrap_policy`

#4037 opened 2026-05-11 06:54 by slocoro

docs: Deadlock when running multi tasks

#4032 opened 2026-05-08 07:05 by xodn348

tests: Gracefully handle missing set_device for mps

#4028 opened 2026-05-04 23:53 by booxter

feat(utils): add max reduction type

#4027 opened 2026-05-03 13:10 by imstevenpmwork

fix(tracker): guard init_trackers and log against None kwargs

#4026 opened 2026-05-02 07:43 by xodn348

[WIP] Make accelerate work end-to-end on AMD ROCm

#4025 opened 2026-04-30 08:16 by Abdennacer-Badaoui

#4023 opened 2026-04-29 07:30 by frozenleaves

#4020 opened 2026-04-24 22:18 by Anai-Guo

fix(checkpointing): save/restore SeedableRandomSampler for map-style datasets

#4019 opened 2026-04-24 16:16 by Anai-Guo

#4017 opened 2026-04-23 10:56 by anishesg

[Checkpointing] Persist DataLoaderShard epoch counter across save/load_state

#4012 opened 2026-04-17 23:08 by s-zx

Security: Unverified remote wheel installation enables supply-chain compromise

#3998 opened 2026-04-06 23:15 by tuanaiseo

Save model-only feature in `save_state`

#3989 opened 2026-03-24 14:36 by 3manifold

Feat: Support dynamic batch size in BatchSamplerShard with even_batches

#3969 opened 2026-03-10 08:24 by yuxinyuan

(fix) remove sampler_is_batch_sampler code in prepare_data_loader(..) wip contributions-welcome

#3469 opened 2025-03-31 13:10 by suzyahyah

handle weight sharing with init_on_device wip

#2737 opened 2024-05-03 01:07 by aws-rhsoln

Enable cpu offload with weights inside the module wip

#2214 opened 2023-12-04 23:11 by SunMarc