strisunshinewentingwang/pytorch

Pull Requests Commits

Add *_only and all/any pytree utilities (#83316)

ezyang committed 2 years ago

b8b54ecc

Add data_dependent_output tag; generalize proxy tensor to test it (#83312)

ezyang committed 2 years ago

d4237226

Don't build nvfuser benchmarks by default (#67857)

peterbell10 committed 2 years ago

d07a9ba1

Fix internal type conversions in floor_divide and trunc_divide (#83288)

peterbell10 committed 2 years ago

42b572bb

[Quant] Make quantizable LSTM scriptable (#83304)

andrewor14 committed 2 years ago

f81b4ae5

[vision hash update] update the pinned vision hash (#82560)

pytorchmergebot committed 2 years ago

d5fb4e2a

[xla hash update] update the pinned xla hash (#83319)

pytorchmergebot committed 2 years ago

86759ad4

Implement QAT for APoT (#83282)

asl3 committed 2 years ago

0e0f8fd0

An improved version of subgraph matcher (#82090)

SherlockNoMad committed 2 years ago

2ca721cd

Temporarily disable test_fx for dynamo (#83307)

huydhn committed 2 years ago

59b1c4e5

Integrate xdoctest - Rebased (#82797)

Erotemic committed 2 years ago

4618371d

fix functionalization <> resnet18, make ProxyTensor work with tensor-less decomps (#83207)

bdhirsh committed 2 years ago

ba90c9f2

[WIP] Add pr sanity check workflow (#83295)

albanD committed 2 years ago

568c6fb9

Revert "Revert "adding a custom caster for c10::SymInt (#82692)"" (#83223)

Krovatkin committed 2 years ago

5b621205

[JIT Test] Add more debugging information for JIT opinfo tests (#83269)

Gamrix committed 2 years ago

99d8eb7b

[JIT SSA] Handle other Tuple outputs (#83222)

Gamrix committed 2 years ago

aac31753

Add fastpath test for mask check flag (#82999)

Yoav Navon committed 2 years ago

dfc97df6

`mul`: sparse-dense + sparse-sparse with 0-dims support take 2. (#82962)

nikitaved committed 2 years ago

b60dc2eb

[ONNX] Bump ONNX Runtime version to 1.12.1 in CI (#81147)

BowenBao committed 2 years ago

bf19e4ea

[ONNX] Update typing and error messages in symbolic_helper (#83007)

justinchuby committed 2 years ago

27108d94

Revert "fix RowwiseMoments vectorization issue on CPU (#81849)"

pytorchmergebot committed 2 years ago

79962017

Revert "RowwiseMoments: use float as acc type for bfloat16 inputs (#81850)"

pytorchmergebot committed 2 years ago

7e6da2fb

Add mac-mps workflow (#83296)

malfet committed 2 years ago

b6156fd4

Register nested matmul as an addition to CompositeImplicit (#82786)

YifanShenSZ committed 2 years ago

7f18ef14

Propagate CUDAOutOfMemoryError to Python. (#83146)

zdevito committed 2 years ago

41287123

[BE] Fix lint in trymerge.py (#83293)

malfet committed 2 years ago

cf6a91b7

[ROCm] Enable bf16-related tests in test_c10d_nccl.py and test_grad_layout_1devicemodule_1replicaperprocess (#82020)

hubertlu-tw committed 2 years ago

cd18b78d

Fix and unskip cpp extension tests for ARM (#83115)

soulitzer committed 2 years ago

b1896255

Set default qengine to QNNPACK on ARM for quantization tests (#83097)

soulitzer committed 2 years ago

95865132

Unskip CompositeCompliance tests for ARM (#83089)

soulitzer committed 2 years ago

ba53efa6

Newer Older