strisunshinewentingwang/pytorch

Pull Requests Commits

Add string versions of argument funcs in jit Node (#45464)

spaugh committed 3 years ago

cdf93b03

Exposes SparseToDenseMask Caffe2 Operator (#45670)

MarcioPorto committed 3 years ago

b234acd4

Add a distributed package reviewer (#45744)

mingzhe09088 committed 3 years ago

ad31068f

Enable type check for torch.quantization.fake_quantize (#45701)

samestep committed 3 years ago

24187a0b

Test torch.svd using complex float and double numbers (#45572)

malfet committed 3 years ago

888f3c12

remove beta defaulting in smooth_l1_loss_backward. added to the bc whitelist (#45588)

bdhirsh committed 3 years ago

4d08930c

some documentation and style fixes to smooth_l1_loss (#45587)

bdhirsh committed 3 years ago

869b2ca0

make broadcasting explanation clearer in matmul doc: #22763 (#45699)

bdhirsh committed 3 years ago

c703602e

VariableKernel calls into scattered C++ api (#44158)

smessmer committed 3 years ago

82cc86b6

Add faithful C++ API (#44087)

smessmer committed 3 years ago

6e2eee2b

Use addmm directly for 1x1 convolution (#45557)

ngimel committed 3 years ago

9201c37d

[quant] PerChannelFloatQParams support for quint4x2 dtype (#45594)

supriyar committed 3 years ago

1a2d3b6a

[quant] creating quint4x2 dtype for quantized tensors (#44678)

supriyar committed 3 years ago

04526a49

Set the default bailout depth to 20 (#45710)

Krovatkin committed 3 years ago

a0d08b21

[docs] Update docs for NegativeBinomial (#45693)

Meghan Lele committed 3 years ago

402caaeb

passing all arguments to sccache wrapper script should be quoted as "$@" (#45582)

jeffdaily committed 3 years ago

36de05db

example of splitting up an FX graph into smaller subgraphs with own submodules (#45404)

Lilyjjo committed 3 years ago

f6dc256b

migrate cuda implementation of take() from TH to ATen (#45430)

bdhirsh committed 3 years ago

1552a926

migrating the take() fn from TH to ATen (#45283)

bdhirsh committed 3 years ago

a015ba8d

Fix the bucketization wrong doc for right argument (#45684)

lixinyu committed 3 years ago

fc4209bd

remove skip annotations since we already disabled the tests wholesale (#45698)

Krovatkin committed 3 years ago

4c1e50eb

win job for the legacy executor (#45612)

Krovatkin committed 3 years ago

cbdba7cc

add an indexer to SymbolicShape (#45450)

Krovatkin committed 3 years ago

0393a1e8

[iOS][CI] Upgrade xcode version to 12.0 (#45677)

xta0 committed 3 years ago

0de5824f

[iOS][CI] Update the dev cert (#45651)

xta0 committed 3 years ago

e8e0fca9

Use CAFFE2_USE_MSVC_STATIC_RUNTIME to determine when to avoid waiting for global destructors on Windows (#43532)

Abaho Katabarwa committed 3 years ago

de3a4801

[reland][quant][graphmode][fx] Merge all quantization mode (#45292) (#45672)

jerryzh168 committed 3 years ago

4f685ecc

Fix BUILD_CAFFE2 if FBGEMM and NNPACK are not built (#45610)

Michael Suo committed 3 years ago

18253f4a

setup: Only include dataclasses for py < 3.8 (#45611)

seemethere committed 3 years ago

5959de3a

[quant] torch.mean add path for unsupported QNNPACK modes (#45533)

z-a-f committed 3 years ago

93be03ce

Older