strisunshinewentingwang/pytorch

Pull Requests Commits

Revert "[inductor] Inline ComputedBuffer computation when there are no reads (#102000)"

pytorchmergebot committed 1 year ago

88961e6d

support ConvBinary in Inductor cpp wrapper (#101393)

chunyuan-w committed 1 year ago

20e6ff37

Revert "[inductor] Handle floordiv and remainder in IndexPropagation (#102277)"

pytorchmergebot committed 1 year ago

f162ab04

Revert "[pt2] add `SymInt` support for `linalg.pinv` (#102367)"

pytorchmergebot committed 1 year ago

da3aba1e

[quant][pt2e] Add Support for DerivedQuantizationSpec (#102282)

jerryzh168 committed 1 year ago

23223402

[inductor] Handle floordiv and remainder in IndexPropagation (#102277)

peterbell10 committed 1 year ago

267a181b

[inductor] Inline ComputedBuffer computation when there are no reads (#102000)

peterbell10 committed 1 year ago

f2dfcb87

[export] Rename graph_module.py to exported_program.py (#102260)

angelayi committed 1 year ago

1e4292a1

[export] ExportedProgram (#102259)

angelayi committed 1 year ago

c4028de4

fix sm86 cuda 21.1 conv threshold issues (#102361)

Fuzzkatt committed 1 year ago

80b916a5

Add dynamo itertools.combinations support (#102379)

wconstab committed 1 year ago

c06d33ce

Replace full_like lowerings with decomps (#101963)

spectrometerHBH committed 1 year ago

76a36159

[aot_autograd][functional_rng] Change calling convention (#102344)

anijain2305 committed 1 year ago

9c4fd72b

s390x simd: disable functions with out-of-bounds reads (#102266)

AlekseiNikiforovIBM committed 1 year ago

bcaa93e8

Merge type stubs torch nn parallel (#102194)

ringohoffman committed 1 year ago

0ed22fce

[Dynamo] Fix if condition on NNModuleVariable (#102335)

yanboliang committed 1 year ago

7b6438da

support ConvUnary in Inductor cpp wrapper (#101392)

chunyuan-w committed 1 year ago

3469f100

[pt2] add `SymInt` support for `linalg.pinv` (#102367)

nkaretnikov committed 1 year ago

0d5b74da

equality assertions (#102256)

avikchaudhuri committed 1 year ago

87510022

[PT2][Quant][BE] Apply formatting to test_quantize_pt2e (#102275)

kimishpatel committed 1 year ago

9b5e4c30

Document faster builds for C++ changes (#102316)

malfet committed 1 year ago

efd774a2

Bump requests from 2.30.0 to 2.31.0 in /tools/build/bazel (#102059)

dependabot[bot] committed 1 year ago

Verified c05a3173

Revert "add additional stream priority for cuda streams (#101956)"

pytorchmergebot committed 1 year ago

6c9b94dc

[MTPG] Enable for some tests in test_fsdp_misc (#102043)

rohan-varma committed 1 year ago

3dfa755a

Add cpp.max_horizontal_fusion_size to control the granularity of horizontal fusion (#99828)

EikanWang committed 1 year ago

ce41faa2

[vision hash update] update the pinned vision hash (#102318)

pytorchupdatebot committed 1 year ago

e1dc793e

[ONNX] Support aten::scatter_reduce (#102048)

titaiwangms committed 1 year ago

fb468b67

Increase mem eff backward performance (#101847)

drisspg committed 1 year ago

ef13fde2

Invoke the bf16 load w/o #elements to bypass the temporary buffer allocation from the performance perspective. (#99822)

EikanWang committed 1 year ago

6f464e0c

Add fast path for BF16 kernel if all the operations within the kernel support bf16 (#99814)

EikanWang committed 1 year ago

c3550d83

Newer Older