strisunshinewentingwang/pytorch

Pull Requests Commits

use new depthwise conv fbgemm interface (#26898)

jspark1105 committed 4 years ago

49266076

Revert D17610292: [pytorch][PR] Choose num_threads in parallel_for based on GRAIN_SIZE

houseroad committed 4 years ago

257b6149

Make TypeDefault, TypeDerived and VariableType anonymous namespaces (#26882)

smessmer committed 4 years ago

092b2f7f

Fix binary size in schema inference (#26878)

smessmer committed 4 years ago

771bcce6

Fix shared_ptr binary size in op registration (#26869)

smessmer committed 4 years ago

54b66c8c

Improve binary size of function schema inference (#26860)

smessmer committed 4 years ago

1a5d641d

Fix c10 registration binary size (#26827)

smessmer committed 4 years ago

84e298e7

use parallel_for in DepthwiseConvKernel (#26879)

ljk53 committed 4 years ago

a6eec839

C++ API parity: TensorTest.Data fix

pbelevich committed 4 years ago

77bfe61f

Make quantized max_pool2d error message more specific and less silly

James Reed committed 4 years ago

388430f6

Support ceil_mode in quantized maxpool

James Reed committed 4 years ago

b1a09dbe

Check if QNNPACK is supported before set (#26935)

IvanKobzarev committed 4 years ago

55fc3778

Set quantized engine backend for mobile in speed_benchmark_torch (#26911)

supriyar committed 4 years ago

8d5c2aa7

Export index_fill and index_copy, fix caffe2 scatter

BowenBao committed 4 years ago

638c4375

batch size 0 tests in BatchMatMul ops (#26874)

jspark1105 committed 4 years ago

d5490c66

batch size 0 support in norm operators (#26894)

jspark1105 committed 4 years ago

ec1f0f08

Migrate lt and lt_ from the TH to Aten (#25998)

ifedan committed 4 years ago

f99bc714

Fix Vec256<T>::abs() for floating point when applied on -0.0 (#26422)

xuhdev committed 4 years ago

9dd8a129

Remove an unused function propagate_names_if_namedtensor_enabled

xuhdev committed 4 years ago

755b7e48

No sccache (#26059)

dncliss committed 4 years ago

ac999365

Use intrinsics for trigonometric functions on CPU (#26431)

xuhdev committed 4 years ago

6f92aa2f

Switch internal CUDA build to C++14 (#26757)

smessmer committed 4 years ago

5c67b014

Fix the Bernoulli distribution sampler (#26864)

EthanSteinberg committed 4 years ago

bf1d957d

Choose num_threads in parallel_for based on GRAIN_SIZE (#26886)

peterbell10 committed 4 years ago

e425bdb8

Get rid of -u (expansion of undefined variable) setting (#26907)

ezyang committed 4 years ago

9f0deb47

fix typo in job name: nigthly->nightly

kostmo committed 4 years ago

b2f671a3

Fix nuclear norm with requires_grad=True (#26303)

vishwakftw committed 4 years ago

43b07ff2

Fix circular deps in loading (#26758)

zdevito committed 4 years ago

0e3389dc

Refactor dispatch structure so fallback code lives inline. (#26367)

ezyang committed 4 years ago

78a52549

Change calling convention of ATenDispatch from getOp to callUnboxed. (#26857)

ezyang committed 4 years ago

272d7c02

Newer Older