strisunshinewentingwang/pytorch

Pull Requests Commits

remove merging label when merge is cancelled (#98967)

PaliC committed 1 year ago

6ebeefb4

add get_device_index for custom device (#98804)

heidongxianhua committed 1 year ago

9a2a6fcf

[inductor] Support integer pow (#88938)

peterbell10 committed 1 year ago

c3186dc8

improvements to torch.gradient docs (#98824)

gusty1g committed 1 year ago

efc90c79

Guard color diagnostics opts by compiler type (#98952)

malfet committed 1 year ago

a8f5d72e

Revert "[export] Constraints API (#98433)"

pytorchmergebot committed 1 year ago

ab761605

[SPMD] Expedite the allreduce call before doing comm_fusion (#98922)

fegin committed 1 year ago

99aacf5c

[Dynamo] Fix Sequential nn module with duplicated submodule (#98880)

yanboliang committed 1 year ago

78ff7ca2

[pt2] add `SymInt` support for `cdist` (#98881)

nkaretnikov committed 1 year ago

8db04e08

[quant][pt2e] Fix a bug in reference quantized module (decomposed mode) (#98903)

jerryzh168 committed 1 year ago

a2e809f2

[AOTAutograd] Fix is-duplicate check in de-dup guard logic (#98932)

awgu committed 1 year ago

3c5a825f

Add shape function for `aten::cross_entropy_loss` (#97875)

vivekkhandelwal1 committed 1 year ago

bb4998b5

Improve symbolic shapes guard logging (#98941)

ezyang committed 1 year ago

5c38c4cf

Revert "[NCCL] Add experimental Nonblocking NCCL Fault Tolerance/Checking (#95715)"

pytorchmergebot committed 1 year ago

1149ba55

[inductor] add `cumprod` to `make_fallback` (#98898)

nkaretnikov committed 1 year ago

c650d7b6

documentation update (#98782)

BarunKGP committed 1 year ago

a38ff4cf

Revert "Move/Fix FakeTensor logic for detecting multiple fake modes (#97186)"

pytorchmergebot committed 1 year ago

48285850

Fix test_mps for macos 13.3 (#98739)

bjhargrave committed 1 year ago

dc52ba29

Make Tensor.__contains__ accept SymInt/Float/Bool. (#98933)

ezyang committed 1 year ago

419ad49e

fix allgather func collective to use maybe_wrap_tensor (#98866)

wanchaol committed 1 year ago

ada7dfff

[spmd] move the param_buffers to the front (#98437)

wanchaol committed 1 year ago

e9954952

Allow set pred with ConstantVariable (#98900)

guangy10 committed 1 year ago

65070e1f

[NCCL] Add experimental Nonblocking NCCL Fault Tolerance/Checking (#95715)

eqy committed 1 year ago

a33eac39

introduce TensorBase::mutable_data_ptr() (#98163)

mikey dagitses committed 1 year ago

09458a2b

[MPS] Add index_fill op (#98694)

qqaatw committed 1 year ago

be8a4eb8

[MPS] Move bitwise ops registration to native_functions.yaml (#98908)

qqaatw committed 1 year ago

01e011b0

[PyTorch] Further reduce cost of TypeMeta::_typeMetaData (by 10x!) (#98105)

swolchok committed 1 year ago

c47464ed

Move/Fix FakeTensor logic for detecting multiple fake modes (#97186)

eellison committed 1 year ago

8a057c44

[dynamo] Remove _dynamo.skip and fold it in _dynamo.disable (#98899)

anijain2305 committed 1 year ago

8654699c

[MPS] Add error inputs check (#98167)

qqaatw committed 1 year ago

71aea7f5

Newer Older