strisunshinewentingwang/pytorch

Pull Requests Commits

Fix test case after get_qparams refactor (#30470)

jerryzh168 committed 4 years ago

fec903ce

Make all optimizers consistent so that they don't change gradients inplace

albanD committed 4 years ago

b0871f21

Change logging to remove the word "error" from info log

Xinyi Zhang committed 4 years ago

45880f42

Specify ordering on singular values and eigenvalues output from torch… (#30389)

vishwakftw committed 4 years ago

dcd9f498

Turn off scalar_check for _th_gather. (#29954)

gchanan committed 4 years ago

dbce53fe

Turn off scalar_checks for torch.take. (#29953)

gchanan committed 4 years ago

72ac4566

Turn off scalar_check for Tensor.set_(Tensor) (#29952)

gchanan committed 4 years ago

79a830af

Export dynamic unbind/split and __getitem__ (#29136)

BowenBao committed 4 years ago

0febff36

Add output_size argument to caffe2 Int8ResizeNearest (#30202)

supriyar committed 4 years ago

2599b9b5

By default ignore RRef leaks during shutdown (#30217)

mrshenli committed 4 years ago

efe1859a

Provide names for operator nodes in ONNX exported graph. (#27342)

Spandan Tiwari committed 4 years ago

06db5ad7

Try exporting ONNX with force_outplace=False (#29466)

BowenBao committed 4 years ago

584be86c

Bug fix: Handle missing keys in observer state dict during load (#30357)

raghuramank100 committed 4 years ago

eccf42fd

Add info about transitive dependencies in case of using local aars (#30128)

IvanKobzarev committed 4 years ago

ab577454

Fix for when PyTorch model trace has RecursiveScriptModules (#30430)

J0Nreynolds committed 4 years ago

085dde59

Add missing std::move (#30411)

smessmer committed 4 years ago

8199596d

Add `get_qparams` and revert the changes to `calculate_qparams` (#30262)

jerryzh168 committed 4 years ago

661a6c8e

Document unsupported types (#30344)

davidriazati committed 4 years ago

46e7f31f

Fix inexistent parameter in document (#24335)

tczhangzhi committed 4 years ago

ab2ec4d8

Refactor QAT Conv module for better extensibility (#30362)

jerryzh168 committed 4 years ago

0b71e7e1

Support to add dequant for each use of Value (#30145)

lly-zero-one committed 4 years ago

b8f50d9c

Improve compare kernel (#29743)

zasdfgbnm committed 4 years ago

25f4ba7c

add default arg for init_method (#30208)

rohan-varma committed 4 years ago

5c6705e6

Add list of supported XCode/CUDA versions to README

smessmer committed 4 years ago

d64e2581

Update osx CI to XCode 9.4 / CUDA 10.0, cudnn 7.6.5 (#30359)

smessmer committed 4 years ago

0517323d

Fix quantized ConvReLU3d test (#30266)

xiaomengy committed 4 years ago

c12f9a12

Stop binding std_single and var_single from TH; they aren't used anymore.

gchanan committed 4 years ago

d7ac90e2

Turn off scalar_check for set_(Storage, ...) (#29950)

gchanan committed 4 years ago

0c673118

Turn off scalar_check for reductions _th_max, _th_min. (#29949)

gchanan committed 4 years ago

71603006

Turn off scalar_check for mode; the underlying code is correct.

gchanan committed 4 years ago

16606e17

Newer Older