strisunshinewentingwang/pytorch

Pull Requests Commits

Disable internal test

kwen2501 committed 2 years ago

8de6204c

Add support to MTPG

kwen2501 committed 2 years ago

f6d3b61c

Add back batch_isend_irecv support for Gloo

kwen2501 committed 2 years ago

bb99f1e9

kwen2501 committed 2 years ago

075c9763

kwen2501 committed 2 years ago

ddc919b0

rebase leftover

kwen2501 committed 2 years ago

d7b913ff

Rethrow exception raised in cm

kwen2501 committed 2 years ago

3a548cb8

kwen2501 committed 2 years ago

c45dd4f7

Add _IllegalWork class

kwen2501 committed 2 years ago

6d7b4538

Fix batch_isend_irecv

kwen2501 committed 2 years ago

dc1afab8

update batch_isend_irecv to use cm

kwen2501 committed 2 years ago

e4f3206d

Move async_op to arg list of cm

kwen2501 committed 2 years ago

10bdd9d6

with ... as cm works

kwen2501 committed 2 years ago

a01b37a8

Address comments

kwen2501 committed 2 years ago

76c94135

Capture op at Python level to reduce pybind calls

kwen2501 committed 2 years ago

9974d6e8

[pt2] add `SymInt` support for fft ops (#99115)

nkaretnikov committed 2 years ago

21681f36

[pt2] add `SymInt` support for `roll` (#99114)

nkaretnikov committed 2 years ago

f89b7c2b

Apply dynamic shapes policy correctly to _base tensor (#99211)

voznesenskym committed 2 years ago

d5f7ec8a

[BE] Update flake8-comprehensions and adapt to rule C418 (#99178)

Skylion007 committed 2 years ago

85f38b8a

make ATen/native/cuda/NLLLoss2d.cu data_ptr-correct (#99179)

mikey dagitses committed 2 years ago

506bd057

make ATen/native/cuda/AdaptiveMaxPooling3d.cu data_ptr-correct (#99185)

mikey dagitses committed 2 years ago

e9201ab6

[CI] Remove inductor skip list for timm_models (#98840)

desertfire committed 2 years ago

34f681c1

[CI] Use expected accuracy csv files to check benchmark test status (#98839)

desertfire committed 2 years ago

a595a506

nn.Linear: dispatch to bsr_dense_mm for half and bfloat16 (#94825)

nikitaved committed 2 years ago

1adb6fa9

make ATen/native/cuda/MaxUnpooling.cu data_ptr-correct (#99189)

mikey dagitses committed 2 years ago

b69f0480

make ATen/native/cuda/UpSampleNearest2d.cu data_ptr-correct (#99186)

mikey dagitses committed 2 years ago

60f914e7

[Dynamo] No graph break for explicit calling Conv{1/2/3}d.forward & ConvTranspose{1/2/3}d.forward (#99015)

yanboliang committed 2 years ago

05809c7d

Enable FSDP ``use_orig_params=True`` mixed precision training when some ranks have no (non-zero sized) parameter shards (#99175)

speediedan committed 2 years ago

157c8690

[dynamo] Support functools.wraps (#98699)

jansel committed 2 years ago

e9be0b0f

[vision hash update] update the pinned vision hash (#99212)

pytorchmergebot committed 2 years ago

b9426ded

Older