strisunshinewentingwang/pytorch

Pull Requests Commits

Add a new op for converting the dense feature to sparse representation

lly-zero-one committed 4 years ago

d6f1346c

Raise error for duplicate params in param group #40967 (#41597)

mariosasko committed 4 years ago

4281240c

[vulkan] Shaders caching (#39384)

IvanKobzarev committed 4 years ago

6367a9d2

Avoid refcount bump in IValue::toStringRef() (#42019)

smessmer committed 4 years ago

d4735ff4

Updates to Scale and Zero Point Gradient Calculation (#42034)

Paul Shao committed 4 years ago

5a6d88d5

Updates to Python Module for Calculation of dX and Addition of Unit Tests (#42033)

Paul Shao committed 4 years ago

c261a894

Renaming variables from dX to dY in Learnable Fake Quantize kernels for Better Clarity (#42032)

Paul Shao committed 4 years ago

e62bf892

Amend docstring and add test for Flatten module (#42084)

alvgaona committed 4 years ago

3e121d96

Remove settings for the logit test case. (#42114)

venkatacrc committed 4 years ago

4290d0be

Added support for Huber Loss (#37599)

alexandrosstergiou committed 4 years ago

11e51749

Enable ProcessGroupGlooTest in CI (take 2) (#42086)

malfet committed 4 years ago

fbdaa555

Grammar Changes (#42076)

rutujak24 committed 4 years ago

96aaa311

DOC: split quantization.rst into smaller pieces (#41321)

mattip committed 4 years ago

b7bda236

DOC: fix two build warnings (#41334)

mattip committed 4 years ago

6af65962

Revert D22741514: [pytorch][PR] Enable ProcessGroupGlooTest in CI

mrshenli committed 4 years ago

47e6d4b3

update cub submodule (#42042)

ngimel committed 4 years ago

b00c05c8

Move qconfig removal into convert() (#41930)

Haixin Liu committed 4 years ago

c5b4f60f

Updates torch.tensor, torch.as_tensor, and sparse ctors to use the device of inputs tensors they're given, by default (#41984)

Mike Ruberry committed 4 years ago

12cd083f

[Resubmit #41318] NCCL backend support for torch bool (#41959)

rohan-varma committed 4 years ago

366c014a

Allow specifying PYTHON executable to build_android (#41927)

malfet committed 4 years ago

38580422

Add prim::EnumName and prim::EnumValue ops (#41965)

gmagogsfm committed 4 years ago

8e03c38a

Remove AllGatherTestWithTimeout (#41945)

osalpekar committed 4 years ago

6287f9ed

Enable ProcessGroupGlooTest in CI (#41985)

malfet committed 4 years ago

45e6f2d6

NCCL must depend on librt (#41978)

malfet committed 4 years ago

cf7e7909

Support aarch32 neon backend for Vec256 (#41267)

kimishpatel committed 4 years ago

dede71d6

caffe2: add PIPELINE tag (#41482)

d4l3k committed 4 years ago

976e6149

update tests to run back-compat check using new binary (#41949)

bradleyhd committed 4 years ago

0c0864c6

Easier english updated tech docs (#42016)

DeepakVelmurugan committed 4 years ago

42a0b51f

updated white list/allow list (#41789)

agniutkarsh committed 4 years ago

becc1b26

.circleci: Make sure to install expect for docs push (#41964)

seemethere committed 4 years ago

7e849132

Older