strisunshinewentingwang/pytorch

Pull Requests Commits

Document `torch.jit.is_tracing()` (#67326)

kmaehashi committed 2 years ago

6ed68f3f

Fix deadlock when forward and backward AD are used at the same time (#67360)

albanD committed 2 years ago

b27b1ff8

Fix indentation in forward_grad.h (#67359)

albanD committed 2 years ago

d3f03af4

[fbcode] Fix operator_benchmark with jit mode (#67382)

dracifer committed 2 years ago

6900aacf

Add test owners for elastic tests (#67293)

janeyx99 committed 2 years ago

eb8b80b7

[LT] Add ir_util for ComputePostOrder (#67282)

desertfire committed 2 years ago

23669480

update coverage ignore to not skip whole modules (#67395)

albanD committed 2 years ago

6293e0ad

[ONNX] Relax check on Prim::PythonOp nodes for ONNX_FALLTHROUGH (#66172) (#67273)

shubhambhokare1 committed 2 years ago

961fd76a

[ONNX] Support conv-bn fusion in blocks (#66152) (#67272)

BowenBao committed 2 years ago

02a78bdb

[ONNX] Use Reciprocal operator instead of Div(1, x). (#65382) (#67271)

garymm committed 2 years ago

9deb6027

fixed type checking errors in fuse.py (#66799)

Onyiee committed 2 years ago

eea20bfa

[SR] NNC out variant for aten::where (#67255)

Mike Iovine committed 2 years ago

7da9c4ed

[quant][embedding qat][bugfix] Fix and test QAT EmbeddingBag from_float error message (#66989)

b-koopman committed 2 years ago

3aadff65

Replace issue templates with new issue forms (#65917)

NicolasHug committed 2 years ago

62feadd7

[Static Runtime][DI] Fuse list unpack and variadic_grouped_accessor_op (#66585)

Mike Iovine committed 2 years ago

6827d36c

specializeGradSumToSize patch - propagate profile_none through profile_ivalue (#63941)

jjsjann123 committed 2 years ago

90b722c5

[sharded_tensor] easier initialization for Shard (#66351)

wanchaol committed 2 years ago

fc664ac2

[sharded_tensor] simplify init_from_local_shards API (#64481)

wanchaol committed 2 years ago

71a67d0c

[quant][graphmode][fx] Add input_idx_to_dtype and ouptut_idx_to_dtype to backend_config_dict (#67067)

jerryzh168 committed 2 years ago

0117ada4

[iOS][CoreML] Remove shape information from TensorSpec (#67412)

xta0 committed 2 years ago

e332d802

[Core ML] Assign Core ML computationUnit to executor (#67411)

xta0 committed 2 years ago

04aba42e

[Core ML] Fix error messages (#67410)

xta0 committed 2 years ago

7e1a53cd

[PyTorch] Reduce refcount bumps in ClassType (#66724)

swolchok committed 2 years ago

fae1c0a4

[PyTorch] Fix extra refcount bumps in ClassAttribute (#66723)

swolchok committed 2 years ago

c8dd90c8

[quant][fx] add pass to duplicate dequant nodes with multi use (#67118)

supriyar committed 2 years ago

1cfdb6f4

Moving python binding to _C and its decl to the right pyi file (#67365)

Krovatkin committed 2 years ago

9e175400

add frozen_numpy to :builtin_registry_cuda target (#67396)

shunting314 committed 2 years ago

882446c1

[SR] Vectorize int version of fmod (#67313)

Hao Lu committed 2 years ago

9ebc6357

[Pytorch Edge] Make some torchbind classes selective (#67340)

JacobSzwejbka committed 2 years ago

dea8b274

[jit] Allow custom class functions to be traced in invokeScriptMethodFromPython(). (#67380)

zhxchen17 committed 2 years ago

f20614af

Newer Older