strisunshinewentingwang/pytorch

Pull Requests Commits

[LTC] Use DataCache in GetIrValueForScalarFromCodegen (#92066)

alanwaketan committed 1 year ago

dfabb916

[BE] meow (#92174)

janeyx99 committed 1 year ago

3debb970

Use binary units for CUDA memory summary (#91854)

milesial committed 1 year ago

421f40e0

Remove unnecessary copies of Scalars for TensorBody template (#92162)

Skylion007 committed 1 year ago

b8057aa1

Move `PyObject` code out of `TensorImpl` into new `PyObjectSlot` class (#92169)

kurtamohler committed 1 year ago

3a0053ab

[torchgen] Add CI job to cover custom ops registration for Executorch (#91291)

larryliu0820 committed 1 year ago

7568484d

Revert "In inductor triton generated code, avoid masking when numel=1 (#91254)"

pytorchmergebot committed 1 year ago

66b324cf

[optim][adadelta] default to foreach when CUDA + differentiable=False (#91896)

janeyx99 committed 1 year ago

d3765509

[PT-D] Fix `send`, `recv` return type (#92152)

awgu committed 1 year ago

cb67d946

[optim] Improve adadelta foreach, group tensors to maximize fast path (#92048)

janeyx99 committed 1 year ago

4af5939d

Apply noexcept to relevant move methods to improve performance (#92156)

Skylion007 committed 1 year ago

3779a75f

Add the new unstable workflow (#92106)

huydhn committed 1 year ago

901a34cc

[GHF] Record how many times PR is revered (#92180)

malfet committed 1 year ago

3794b464

[ROCM] Modify transcoding: absolute path ->relative path (#91845)

xiabo123 committed 1 year ago

70b3ea59

MYPYNOFOLLOW for test_utils (#92136)

clee2000 committed 1 year ago

214c0fdc

[CI][ROCm] skip multiprocessing tests that trigger hangs (#92101)

jeffdaily committed 1 year ago

04689ae2

[cuDNN][cuDNN V8 API] Always build assuming cuDNN >= 8.0 (#91527)

eqy committed 1 year ago

4d07ad74

Revert "Pytorch-bot test (#92163)"

pytorchmergebot committed 1 year ago

4d269037

Pytorch-bot test (#92163)

clee2000 committed 1 year ago

7fe3c64b

Call profiler step via optimizer post hook (#90101)

anupambhatnagar committed 1 year ago

f4b804ee

Update CMakeLists.txt since MacOS linker doesn't support whole-archive (#91736)

TachikakaMin committed 1 year ago

6783db13

[follow-up] Python Attr Serialization (#88913)

kshitij12345 committed 1 year ago

745fe35d

Do not leak SkipFrame exception to parent frames (#91059)

BowenBao committed 1 year ago

a72bcb33

add docstring for adam differentiable parameter (#91881)

nouranali committed 1 year ago

a60125e2

[BE] Use nested namespaces in .cpp/.cu files (#92100)

malfet committed 1 year ago

8f1c3c68

Fix torch.where signature mismatch that was caused by torchgen (#91627)

LucaLumetti committed 1 year ago

a4a0195c

[torchdim] Fix Python 3.11 bytecode decoding in dims (#91290)

zdevito committed 1 year ago

accecd7b

Update sgd doc to insist on momentum buffer initial value (#92111)

albanD committed 1 year ago

60e37a6e

Improve type hints for Module forward hooks (#92061)

ringohoffman committed 1 year ago

a26e5e21

Add kwargs support to torch.export() API (#92013)

Thiago Crepaldi committed 1 year ago

890b6828

Newer Older