strisunshinewentingwang/pytorch

Pull Requests Commits

Fix deadlock in ExecutionTraceObserver (#119242)

shengfukevin committed 230 days ago

6fc775ae

Refactor Symint Deduping to separate pass (#118938)

eellison committed 230 days ago

d0ca849f

Revert "Add meta registration for _foreach_norm (#118604)"

pytorchmergebot committed 230 days ago

dea15c9f

[quant][pt2e] Allow users to override train/eval behavior (#119091)

andrewor14 committed 230 days ago

6c1cca15

Revert "[c10d] PGNCCL refactor part 1: adds assert size==1 (#119099)"

pytorchmergebot committed 230 days ago

9d46fe60

Make filename optional in update_failures.py (#119289)

jansel committed 230 days ago

0f68bcaa

Change PrivateUse1's resize_bytes to PrivateUse1HooksInterface (#117839)

CLiqing committed 230 days ago

422b4271

[dynamo] refactor CacheEntry and ExtraState to eval_frame.c to C++ (#118438)

williamwen42 committed 230 days ago

ae4e866b

[fix] accounting for dilation in pool padding assertion (#118897)

galqiwi committed 230 days ago

73f0fdea

[dynamo] Skip dynamo when inside a functorch context (#118901)

jansel committed 230 days ago

ec31d115

Update auto_functionalize docs (#119228)

angelayi committed 230 days ago

f3645fc3

Migrate last lbfgs test over to OptimizerInfo (#119283)

janeyx99 committed 230 days ago

f85b0ea8

Introduce size oblivious guards (#118579)

ezyang committed 230 days ago

3f0fd368

[dynamo] support comparing stream with constant (#119199)

ydwu4 committed 230 days ago

5410385c

[mypy] declare type for DynamoTestCase._exit_stack (#119084)

ColinPeppler committed 230 days ago

fa157af6

Add a short code snippet in the RNN doc (#119150)

lancerts committed 230 days ago

238d87f7

Move catch_errors_wrapper to convert_frame (#119253)

ezyang committed 230 days ago

169c0700

Make start compiling stack trace omit framework frames (#119251)

ezyang committed 230 days ago

790858af

Reserve sizes in c10::VaryingShape::concrete_sizes(), c10::TensorType::computeStrideProps() (#119189)

IosifSpulber committed 230 days ago

22669843

[Dynamo] Remove build_checkpoint_variable from call_getattr (#119236)

yanboliang committed 230 days ago

8ee9f26c

Add torch.backends.mha.get_fastpath_enabled to FUNC_INLINELIST (#118979)

ckluk2 committed 230 days ago

2ad3599a

Bugfix to MixtureSameFamily's _pad_mixture_dimension (#118947)

CJMenart committed 230 days ago

a77be631

Revert "Add FakeTensor support to torch._utils._rebuild_tensor (#108186)"

pytorchmergebot committed 230 days ago

499040ac

[decomp] Add tests for different dtypes to SDPA decomposition (#119239)

larryliu0820 committed 230 days ago

1e4b408b

Update scatter_reduce_ test with parallel backend check (#118708)

leslie-fang-intel committed 230 days ago

85033759

[inductor] Support sympy.expr in user-defined Triton kernel grid fn (#119165)

ColinPeppler committed 230 days ago

7d7a3f0b

Fix type hints on nn.attention.sdpa_kernel (#119140)

ajbrent committed 230 days ago

8a8e7047

[CPU] Optimize softmax as flash attention v2 (#118957)

Valentine233 committed 230 days ago

720f7811

[c10d] PGNCCL refactor part 1: adds assert size==1 (#119099)

kwen2501 committed 230 days ago

4ab852b6

[inductor] Implementing missing magic methods on IR values. (#118933)

amjames committed 230 days ago

884b6d2a

Newer Older