strisunshinewentingwang/pytorch

Pull Requests Commits

OpInfo: specifying sparse sample input function implies the corresponding layout support (#100392)

pearu committed 1 year ago

6b2ecb12

Fix sum OpInfo for sparse sample inputs and assert coverage for sparse-enabled operators (#100391)

pearu committed 1 year ago

3ae0e23b

Move tracked nn_modules from OutputGraph to TracingContext (#100457)

voznesenskym committed 1 year ago

ffcbd1c2

Remove special casing for stride/size setup for guards (#100456)

voznesenskym committed 1 year ago

2439090b

Avoid using einsum for torch.cat DTensor propogation (#100251)

mrshenli committed 1 year ago

9439cb0e

[ONNX] Add RemoveConstantInputStep to adapt torch inputs to ONNX inputs (#100252)

titaiwangms committed 1 year ago

d23dbfff

[inductor] Change the default value of layout (#100254)

desertfire committed 1 year ago

6b5f5000

Revert "[WIP] enable cuda graphs support for flash attention with dropout (#100196)"

pytorchmergebot committed 1 year ago

c3aa59c8

Fix hosts update for binary build (#100507)

malfet committed 1 year ago

dc4a2531

Remove ci: sev label and details from ci-sev.md tempalte (#100504)

kit1980 committed 1 year ago

e4ad67f9

Revert "[inductor] Cleanup strip_last_size logic (#100305)"

pytorchmergebot committed 1 year ago

34e90b8d

[PT2E][Quant] Fix but in quant spec of symmetric static quant (#99398)

kimishpatel committed 1 year ago

8ec0a939

Handle trailing masked column behavior for nested tensor (#100113)

mikekgfb committed 1 year ago

8430430e

[dashboard] higher tolerance for AlbertForQuestionAnswering (#100277)

anijain2305 committed 1 year ago

0acfe2ce

[inductor] Cleanup strip_last_size logic (#100305)

jansel committed 1 year ago

de7793d5

[WIP] enable cuda graphs support for flash attention with dropout (#100196)

ngimel committed 1 year ago

32615618

[CI] Change the dashboard run to once a day (#100499)

desertfire committed 1 year ago

a587f1ff

Populate download.pytorch.org IP to container (#100475)

malfet committed 1 year ago

7ff71a3a

Revert "PyTorch -> C++17 (#98209)" (#100497)

clee2000 committed 1 year ago

2ec6eb3d

Revert "Migrate jobs from windows.4xlarge windows.4xlarge.nonephemeral instances (#100377)"

pytorchmergebot committed 1 year ago

543b7ebb

Migrate jobs from windows.4xlarge windows.4xlarge.nonephemeral instances (#100377)

jeanschmidt committed 1 year ago

7caac545

Move pattern match for foreach before bulky if-else in `save_variables` (#100445)

crcrpar committed 1 year ago

311c2bb7

Revert "Mount /etc/hosts into container (#100475)"

pytorchmergebot committed 1 year ago

e8a1d0be

[dynamo][moco] Disallow_in_graph distributed APIs (#100071)

anijain2305 committed 1 year ago

5fbb4066

[c10d] Add new Store methods: append, multi_get, multi_set. (#100379)

Rodrigo Kumpera committed 1 year ago

0dc671c2

PyTorch -> C++17 (#98209)

r-barnes committed 1 year ago

8f0c825d

ci: win cpu test -> trunk, cuda test -> periodic (#100478)

seemethere committed 1 year ago

50b0fff0

[BE] Update cutlass with NVIDIA upstream changes to 3.1 (#100333)

Skylion007 committed 1 year ago

0efab604

enable headdims > 64 for flash attention on sm90 (#99776)

Fuzzkatt committed 1 year ago

06bf5d4d

[pt2] add `SymInt` support for `dsplit`, `hsplit`, `vsplit` (#100352)

nkaretnikov committed 1 year ago

279f3cd0

Newer Older