strisunshinewentingwang/pytorch

Pull Requests Commits

[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

[ROCm] Add sparse mappings for CUDA->HIP translation (#67323)

IvanYashchuk committed 2 years ago

2267a984

Update extending doc to cover forward mode AD (#66962)

albanD committed 2 years ago

708f7b12

[ONNX] Add dim argument to all symbolic (#66093) (#67270)

shubhambhokare1 committed 2 years ago

d9a56689

[ONNX] Update onnxruntime to 1.9 for CI (#65029) (#67269)

garymm committed 2 years ago

cb15df76

Fix sign warnings in CUDA kernels (#66753)

r-barnes committed 2 years ago

99003101

Add dummy bfloat16 VSX implementation (#67331)

malfet committed 2 years ago

3a1aa31a

Wrap TRTInterpreter result with wrapper (#67307)

Shirong Wu committed 2 years ago

7484941e

Set kernel func name from AOT Compiler (#67229)

priyaramani committed 2 years ago

fa70d72e

Set test owners for onnx tests (#66860)

janeyx99 committed 2 years ago

5347dab8

[Static Runtime][DI] Add variadic grouped_accessor_op (#66289)

Mike Iovine committed 2 years ago

72e25c9f

Add fp16/fp32 autocasting to JIT/TorchScript (#63939)

jjsjann123 committed 2 years ago

1ec732bc

[skip-ci] .github: Set linux gpu instances to be non-ephemeral (#67345)

seemethere committed 2 years ago

0101b1ea

[jit] Remove graph() call from abstract Function interface. (#65967)

zhxchen17 committed 2 years ago

b55a2500

Sparse CSR CUDA: add `triangular_solve_out` (#61858)

IvanYashchuk committed 2 years ago

7c48b9ee

[fx]Early return if a node tries prepend self (#67068)

842974287 committed 2 years ago

4b9464f4

Revert D31945507: .github: Switch 8xlarge to 4xlarge instance_type

seemethere committed 2 years ago

2669e4ed

GHA: add back runner type for distributed tests (#67336)

janeyx99 committed 2 years ago

7d1c0992

Add support for vmap+fwdAD for basic out-of-place op (#66291)

soulitzer committed 2 years ago

f2f7b02b

Add barrier to ProcessGroup trampoline (#67236)

mrshenli committed 2 years ago

a3aa9df5

[tensorexpr][ir][quant] Adding qscale and qzero to tensorexpr IR Buf (#66675)

IvanKobzarev committed 2 years ago

e52d0e77

Enable c10d trampoline tests on MacOS (#67205)

mrshenli committed 2 years ago

632719c2

[hpc][inference] enable cuda graph in engine holder (#66738)

Bangsheng Tang committed 2 years ago

c88da701

Newer Older