strisunshinewentingwang/pytorch

Pull Requests Commits

[BE] Cleanup where_self_out (#121494)

malfet committed 196 days ago

9cc89970

Enable `out` OpInfo testing for `torch.where` (#121473)

malfet committed 196 days ago

1866ee67

Update Quantizable LSTM to support QAT (#121448)

Ninja91 committed 196 days ago

0dd21c0c

Deprecate torch.autograd.function.traceable, is_traceable (#121413)

zou3519 committed 196 days ago

b52e0bf1

[tp] remove deprecated tp_mesh_dim arg (#121432)

wanchaol committed 196 days ago

08460f4b

[tp] doc fixes (#121431)

wanchaol committed 196 days ago

30982ce0

Fix round robin sharding (#121022)

clee2000 committed 196 days ago

effdea5f

Update XLA pin (#121501)

malfet committed 196 days ago

9d83f9dc

[AOTDispatch] Return mutated inputs directly when keeping mutations (#120514)

peterbell10 committed 196 days ago

a2a8c1fd

[DTensor][XLA] support XLA backend in distirbute_module API (#121355)

yeounoh committed 196 days ago

f7ec984b

Batch Norm Consolidation (#116092)

andrewor14 committed 196 days ago

7b4f70ed

Add links to _ex variants in all linalg functions that support them (#121451)

lezcano committed 196 days ago

c253d1c1

[Quant] Add the operator of decomposed fake quant per channel (#121297)

leslie-fang-intel committed 196 days ago

975d4284

Update link to OpenVINO backend in torch.compiler.rst (#121303)

kblaszczak-intel committed 196 days ago

8ed09321

fix accidental specialization with faketensor input checks (#121460)

avikchaudhuri committed 196 days ago

b3f24b57

[DCP][state_dict][doc] Update the distributed state_dict document (#121290)

fegin committed 196 days ago

2e789ad5

suggested fixes for congruences (#121418)

avikchaudhuri committed 196 days ago

e628f2cc

[DCP] Makes async_save public (#121325)

LucasLLC committed 196 days ago

96ed37ac

[DCP][state_dict][doc] Fix the documents for distributed_state_dict (#121276)

fegin committed 196 days ago

13366a10

[inductor] Make some improvements to FX graph caching (#117888)

masnesral committed 196 days ago

72dd9b24

[DCP] Removes `no_dist` and `coordinator_rank` from public DCP API's (#121317)

LucasLLC committed 196 days ago

909d73d8

more passing dynamo tests (#121378)

aorenste committed 196 days ago

23ac0cd5

[nit][DCP][DSD] Remove Unused Variables in test_state_dict.py (#121204)

wz337 committed 196 days ago

4186c365

Revert "[fake_impls] Fix seed/offset device for attention kernels (#120839)" (#121447)

davidberard98 committed 196 days ago

0f8c9acc

[dtensor][TP] check funcol calls and improve doc for loss parallel (#121366)

tianyu-l committed 196 days ago

dc514b96

[fx] Preserve Fx graph node order in partitioner across runs (#115621)

kareemshaik80 committed 196 days ago

25c74a93

make dyanmo work with _LazyGraphModule.lazy_forward (#121259)

shunting314 committed 196 days ago

7dc1ab89

[Torch Elastic][Draft] Refactor SubprocessHandler to separate module for easier subclass (#120373)

cniii committed 196 days ago

9bff1599

[dynamo][guards-cpp-refactor] Func defaults and kwdefaults accessor (#121338)

anijain2305 committed 196 days ago

c86a1ce1

[dynamo][guards-cpp-refactor] Permit dict version guard in DictGuardManager (#121327)

anijain2305 committed 196 days ago

79a04f2d

Newer Older