strisunshinewentingwang/pytorch

Pull Requests Commits

fix normal with empty std (#66463)

ngimel committed 2 years ago

122c6ee7

Use shared CUPTI by default (#65401)

ezyang committed 2 years ago

90677145

Back out "Revert D30710710: [Pytorch Edge] Support profiling kineto events from external source" (#66421)

kimishpatel committed 2 years ago

f5fc754a

[jit] Eliminate malloc & recursive refcount bumps in HashType::operator() (#65348)

swolchok committed 2 years ago

fb66cdb1

[jit] Compare object identity first in ClassType::operator== (#65347)

swolchok committed 2 years ago

abdaca97

remove is_reference from all is_output_quantized (#66456)

rahxephon89 committed 2 years ago

7d050ba5

[fx2trt] open source tests for acc tracer (#66302)

jerryzh168 committed 2 years ago

be17bf11

Upgrade onnx submodule to 85546f8c44e627f8ff1181725d03cc49f675e44f (#66427)

houseroad committed 2 years ago

0d8b1744

[FX] fuse permute021 linear pass for trt lowering (#66362)

Yinghai Lu committed 2 years ago

7b2190f1

Fix lstsq to work with inputs that require grad (#66426)

IvanYashchuk committed 2 years ago

6cbe7125

Remove native_functions.yaml dependency from PointwiseOps (#64172)

peterbell10 committed 2 years ago

165c65b1

Remove native_functions.yaml dependency from PowKernel (#64171)

peterbell10 committed 2 years ago

e646edc0

Remove native_functions.yaml dependency from unary ops (#64170)

peterbell10 committed 2 years ago

ec9636eb

[DataPipe] Fix HttpReader IterDataPipe Issue with streaming (#66432)

NivekT committed 2 years ago

52ba8b3a

[TensorExpr] Fix lowering for aten::t. (#65859)

Mikhail Zolotukhin committed 2 years ago

50e71662

[TensorExpr] Fix lowerings for aten::view and aten::reshape. (#65852)

Mikhail Zolotukhin committed 2 years ago

9fafeb15

[TensorExpr] Use schema instead of op name in NNC lowerings. (#65843)

Mikhail Zolotukhin committed 2 years ago

54cc93f4

[TensorExpr] Nuke TE shape inference. (#65554)

Mikhail Zolotukhin committed 2 years ago

c99a3f26

[Pytorch Edge] Improve bundled inputs name error handling (#65856)

JacobSzwejbka committed 2 years ago

778ea987

Revert D31386275: Migrate THCState to ATen

suo committed 2 years ago

68d4c180

[PyTorch] Split c10 Type.cpp into two files to allow targets to include one of them (#66445)

larryliu0820 committed 2 years ago

ced5ff93

[deploy] fix oss build (#66347)

suo committed 2 years ago

d541d100

Fix batch_isend_irecv tests for err case (#63112)

zasdfgbnm committed 2 years ago

80323e3a

[FX] Fix issue where GraphModule.delete_all_unused_submodules deletes submodules from called leaf modules (#66430)

James Reed committed 2 years ago

45fd68ac

Migrate THCState to ATen (#65948)

peterbell10 committed 2 years ago

479dc915

Call `PyArray_Check` only if NumPy is available (#66433)

malfet committed 2 years ago

a72ac5d1

Quantization docs: add pages for Numeric Suite (Eager and FX) (#66380)

vkuzo committed 2 years ago

a515b4d2

Improve quantization API docs (#66379)

vkuzo committed 2 years ago

16da541c

Fix type checking errors in torch/quantization/fx/qconfig_utils.py (#66428)

EshikaShah committed 2 years ago

f3600e9d

Back out "Revert D31495086: open source engine_layer_visualize.py" (#66431)

jerryzh168 committed 2 years ago

17505806

Newer Older