huggingface/accelerate

Pull Requests Commits

Release: v1.11.0

SunMarc committed 205 days ago

9a81156b

refactor: nit change for get_parameters_from_modules (code debt) (#3815)

kmehant committed 211 days ago

Verified 5998f862

Fix tracking swanlab (#3810)

SunMarc committed 215 days ago

Verified f0313a64

Bump to python3.10 + update linter (#3809)

SunMarc committed 215 days ago

Verified df0c1870

SunMarc committed 217 days ago

Verified bc2478a4

fix (skip) cache flush when original device is `cpu` and offloaded to disk `meta` (#3796)

Qubitium committed 217 days ago

Verified 057edec2

Remove deprecated FindTiedParametersResult (#3786)

cyyever committed 236 days ago

Verified 14383311

Revert "fix: correct dictionary unpacking in recursively_apply function (#3766)" (#3787)

SunMarc committed 236 days ago

Verified a737437c

rm mlflow (#3783)

SunMarc committed 236 days ago

Verified 6997855a

Add optional typing (#3769)

cyyever committed 237 days ago

Verified 401075ff

refactor: Use `with` in Accelerator.autocast()instead of ` __enter__()` and` __exit__()` for more elegant style. (#3767)

EquationWalker committed 237 days ago

Verified 8031e24e

fix: correct dictionary unpacking in recursively_apply function (#3766)

qgallouedec committed 237 days ago

Verified 3db9fb69

switch XPU ccl backend to torch-builtin xccl in test_zero3_integration (#3773)

yao-matrix committed 237 days ago

Verified fe795fd3

Lower complexity of get_balanced_memory by adding a set (#3776)

SamuelBarryCS committed 238 days ago

Verified 409b356f

enable 2 model hook ut cases on XPU (#3774)

yao-matrix committed 239 days ago

Verified 1b50d939

Added Tip for better rendering (#3781)

sergiopaniego committed 240 days ago

Verified e79f3836

fix Muti node CUDA error: invalid device ordinal #3775 (#3779)

RicardoDominguez committed 242 days ago

Verified 0cb1a334

use reset_peak_memory_stats on xpu (#3772)

yao-matrix committed 243 days ago

Verified dfdc2190

fix FSDP2 test case failure on XPU (#3771)

yao-matrix committed 243 days ago

Verified 45959d7b

Fix: typo makes tests fail (#3765)

S1ro1 committed 246 days ago

Verified 8b493524

fix: torch_npu import error (#3764)

yanyongyu committed 246 days ago

Verified 9ead94e5

feat: allow mixed precision policy as dtype (#3751)

kmehant committed 247 days ago

Verified a0bc36e8

Fix typos (#3753)

cyyever committed 247 days ago

Verified 8830e58a

make torch_native_parallelism examples device agnostic (#3759)

yao-matrix committed 247 days ago

Verified 40ebb4be

fix: model.set_requires_gradient_sync(False) should be called to turn off gradient synchronization in FSDP2 (#3762)

EquationWalker committed 249 days ago

Verified ec92b1af

CP docs typos fixed (#3761)

sergiopaniego committed 250 days ago

Verified 62ede1ed

fix: specify device for process_tensor in example usage (#3755)

qgallouedec committed 252 days ago

Verified 9f9c490c

xpu INT64 all_gather issue fixed in 2.9 (#3756)

yao-matrix committed 252 days ago

Verified 8b55e62b

Add bf16/fp16 support for amp with mps device (#3373)

SunMarc committed 258 days ago

Verified 0e4419b3

Add support for TE MXFP8 recipe in accelerate (#3688)

pstjohn committed 259 days ago

Verified 3b67c216

Older