strisunshinewentingwang/pytorch

Pull Requests Commits

[quant] Fix for the conv1d kernel shape (#36397)

z-a-f committed 4 years ago

d83509e6

Disables complex clamp (#36373)

Mike Ruberry committed 4 years ago

0c9bf649

Adds NumPy array x Torch tensor binary ufunc interaction test (#35945)

Mike Ruberry committed 4 years ago

254be6a2

[ONNX] Enable constant folding for Shape (#35386)

KsenijaS committed 4 years ago

4f728c9d

[TensorExpr] Add new file bounds_inference.cpp to BUILD.bazel. (#36440)

Mikhail Zolotukhin committed 4 years ago

e3af0c9f

Enable building of FakeLowP ops (#36170)

yinghai committed 4 years ago

c1efe1dd

Disables complex min and max (#36377)

Mike Ruberry committed 4 years ago

7aa6a8fd

Improve boxed dispatch performance (#33313)

smessmer committed 4 years ago

7b9ab916

Remove functor factories in KernelFunction (#35488)

smessmer committed 4 years ago

22212a82

[Lite Interpreter] Move Implicic ops to register_prim_ops.cpp (#36406)

iseeyuan committed 4 years ago

91441ae8

[TensorExpr] Implement LoopNest::computeAt (#36112)

Mikhail Zolotukhin committed 4 years ago

df5f0a04

[TensorExpr] Bounds inference (#35120)

Mikhail Zolotukhin committed 4 years ago

397aa46a

Move unboxing to after dispatch for ops with manual kernel registrations (#36398)

smessmer committed 4 years ago

c856a2cb

Fix view_complex_as_float for empty tensors (#36415)

pbelevich committed 4 years ago

7e8c27ed

Revert D20961711: [pytorch][PR] Returns float tensors for complex inputs to abs

houseroad committed 4 years ago

742c7797

[DistAutograd x JIT] Capture global state, dist autograd current context id, before thread switching triggered by JIT future.wait() (#36395)

xush6528 committed 4 years ago

ae452a81

Revert D20931186: Enable c10 unboxing for ops with TensorList

houseroad committed 4 years ago

0dbb21f8

Updating submodules

svcscm committed 4 years ago

409346ee

Catch exception in distributed engine callbacks. (#36118)

pritamdamania committed 4 years ago

5b331e86

Updating submodules

svcscm committed 4 years ago

d71aeece

Upstream generic device test patch. (#36321)

Ailing Zhang committed 4 years ago

e8923989

Remove experimental c10 ops (#36394)

smessmer committed 4 years ago

4305c7f9

Move fakelowp tests from glow to caffe2 (#36409)

Yinghai Lu committed 4 years ago

6920b135

Revert D20958928: [pytorch][PR] Port masked_select cuda from TH to ATen

Mike Ruberry committed 4 years ago

bd476112

Revert D20523080: [pytorch] reduce memory footprint in fused conv QAT ops

Mike Ruberry committed 4 years ago

86e8c49f

hardswish: add cuda kernels (#36350)

vkuzo committed 4 years ago

eddbee19

[caffe2] Use cpuinfo in perfkernels to simplify build dependency (#36371)

Dmytro Dzhulgakov committed 4 years ago

7576cf8d

Port masked_select cuda from TH to ATen (#35429)

kurtamohler committed 4 years ago

343f2c09

Open source the missing part of FakeFp16 ops (#36353)

Yinghai Lu committed 4 years ago

d27dccfd

Updating submodules

svcscm committed 4 years ago

c029aaa2

Newer Older