strisunshinewentingwang/pytorch

Pull Requests Commits

Revert "Simple Custom Operator API, V0 (#98440)"

pytorchmergebot committed 1 year ago

f497031d

Revert "Reland python ops (#99170)"

pytorchmergebot committed 1 year ago

1c042a21

Fix the pt2e UT path after refactor (#99402)

leslie-fang-intel committed 1 year ago

c97dd8e1

[dynamo 3.11] implement 3.11 exceptiontable (#96511)

williamwen42 committed 1 year ago

88c8c2b7

Python binding to set/get CUDA rng state offset (#98965)

anijain2305 committed 1 year ago

8214fe07

Mark doctr_det_predictor as broken on master (#99370)

ezyang committed 1 year ago

b290381e

[CUBLAS] Specify alignment for `cuBlasLt` `addmm` (#98975)

eqy committed 1 year ago

5c5ad535

Fix bug in check required output size in _as_strided_scatter_meta (#98483)

lantiankaikai committed 1 year ago

5b692fd8

[Inductor] Unify Inductor CUDA & CPUT unit tests input clone function (#99118)

yanboliang committed 1 year ago

2611fccf

[BE][DTensor] fix DTensor equal op (#99014)

XilunWu committed 1 year ago

964c7e3e

Test and document dynamo backward hooks support (#99382)

wconstab committed 1 year ago

e6aa8e07

[JIT] clarify errors due to non-literal indexing into ModuleList, ModuleDict (#98606)

davidberard98 committed 1 year ago

cde35b40

[CI] Move last iOS job to periodic (#99388)

malfet committed 1 year ago

a763d948

[CI] Update torchbench pin (#99386)

malfet committed 1 year ago

4ffd407d

Switch calling convention back to real tensors (#99320)

ezyang committed 1 year ago

780922c2

Delete use_functionalize feature flag (#99317)

ezyang committed 1 year ago

a109453d

Delete functorch use_fake_tensor and debug_fake_cross_ref (#99314)

ezyang committed 1 year ago

17d7be68

Make run_fwd_maybe_bwd work with int inputs (#99365)

ezyang committed 1 year ago

2471eac6

[DataLoader] Add missing documentation for arg in DataLoader (#99371)

NivekT committed 1 year ago

3d8498f9

[ONNX] Retire 'DynamoOptimizeExporter' (#99202)

BowenBao committed 1 year ago

436edc5a

[inductor][easy] create a wrap for triton do_bench function (#99216)

shunting314 committed 1 year ago

694ed70e

Skip sccache initialization on MacOS (#99121)

huydhn committed 1 year ago

063173cb

Rename sym_shapes logger to dynamic (#99335)

ezyang committed 1 year ago

6df87b2e

Revert "Skip sccache initialization on MacOS (#99121)"

pytorchmergebot committed 1 year ago

6212a85a

enable data type propagation (#98065)

zhuhaozhe committed 1 year ago

59e343b1

[FSDP][optim_state_dict][Easy] Temporarily disable rank0_only=True for use_orig_paramscaseEas (#99354)

fegin committed 1 year ago

01fdcbdc

Revert "Revert "Expandable blocks in allocator (#96995)"" (#99275)

zdevito committed 1 year ago

7ff1f3f3

fix typo in gen_functionalization_type.py (#99303)

eltociear committed 1 year ago

99c6d46c

Updates NCCL to 2.17.1 (#97843)

syed-ahmed committed 1 year ago

b003000f

Skip sccache initialization on MacOS (#99121)

huydhn committed 1 year ago

c2fd198c

Newer Older