Skip to content

Pull requests: huggingface/accelerate

New pull request New
16 Open 1,705 Closed
16 Open 1,705 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

add support for port 0 auto-selection in multi-GPU environments
#3501 opened Apr 11, 2025 by hellobiondi Loading…
3 of 5 tasks
1
WIP: Compose TP + DDP/FSDP2
#3498 opened Apr 10, 2025 by S1ro1 • Draft
2
Fix duplicates created when using PartialState.split_between_processes apply_padding
#3490 opened Apr 7, 2025 by zcakzhuu Loading…
5 tasks done
1
3
Fsdp2 cpu ram efficient loading
#3482 opened Apr 4, 2025 by S1ro1 Loading…
1
(fix) remove sampler_is_batch_sampler code in prepare_data_loader(..)
#3469 opened Mar 31, 2025 by suzyahyah Loading…
8
check that an object has .float() method
#3451 opened Mar 20, 2025 by eljandoubi Loading…
5
fix "clear_device_cache" in hooks.py causing a CUDA runtime error in multiple GPUs environment.
#3430 opened Mar 9, 2025 by CyberVy Loading…
8
[Feat] Layerwise casting hook
#3427 opened Mar 7, 2025 by sayakpaul Loading…
2 of 3 tasks
7
Support TP + FSDPv2 / HSDP or just FSDPv2 / HSDP
#3395 opened Feb 13, 2025 by kmehant • Draft
5 tasks
4
Add bf16/fp16 support for amp with mps device wip Work in progress
#3373 opened Jan 29, 2025 by SunMarc • Draft
3
[WIP] optimize infer_auto_device_map for multi-GPU allocation wip Work in progress
#3321 opened Jan 2, 2025 by Nech-C • Draft
3 of 5 tasks
4
[FSDP] Make env var and dataclass flag consistent for cpu_ram_efficient_loading
#3307 opened Dec 19, 2024 by SumanthRH Loading…
1 of 5 tasks
8
Some adjustment for supporting Deepspeed-Ulysses wip Work in progress
#2877 opened Jun 20, 2024 by zeyugao Loading…
1 of 5 tasks
17
handle weight sharing with init_on_device wip Work in progress
#2737 opened May 3, 2024 by aws-rhsoln Loading…
5 tasks
7
[DO NOT MERGE] add all level buffer support when computing infer_auto_device_map enhancement New feature or request feature request Request for a new feature to be added to Accelerate wip Work in progress
#2663 opened Apr 12, 2024 by SunMarc Loading…
3
Enable cpu offload with weights inside the module wip Work in progress
#2214 opened Dec 4, 2023 by SunMarc • Draft
10
ProTip! Adding no:label will show everything without a label.