strisunshinewentingwang/pytorch

Pull Requests Commits

Make init_method url appending step re-usable by both init_process_group and init_model_parallel(init_rpc) (#28226)

xush6528 committed 4 years ago

59402f51

Make RRef::LocalValue return Future (#28025)

mrshenli committed 4 years ago

e31adeb4

Make RRef::toHere() return a jit::Future (#27943)

mrshenli committed 4 years ago

58873776

Updating submodules

svcscm committed 4 years ago

9c345473

static initialization order with mutex (#28243)

EscapeZero committed 4 years ago

61d40b80

workaround for raw string bug in VS2019 (#28349)

Mingbo Wan committed 4 years ago

80083223

Scripts for setting up benchmark projects (#28469)

xta0 committed 4 years ago

896b5d91

Ignore F401 in all __init__.py without putting noqa (#25823)

narumiruna committed 4 years ago

d83389d3

export group_norm (#27071)

neginraoof committed 4 years ago

76d262d4

fix the document of kaiming initialization (#25638)

robert780612 committed 4 years ago

d081de67

fix docs for lr (#28026)

vincentqb committed 4 years ago

cbddc77a

is_set_to: unify TH/THC implmentation and genericize test_is_set_to. (#28422)

gchanan committed 4 years ago

bee4aca2

Change activation modules in C++ from using Tensor& to Tensor (#28501)

t-vi committed 4 years ago

09ad464d

Fixed behavior of div_factor parameter in optim.lr_scheduler.OneCycleLR (#28217)

Timothy Man committed 4 years ago

1c53a74e

Updating submodules

svcscm committed 4 years ago

76c70559

Return 0-numel empty tensor from symeig when eigenvectors=False (#28338)

vishwakftw committed 4 years ago

657430e1

Add multiplicative lr. (#27254)

vincentqb committed 4 years ago

e4f40bf3

Correct math formatting for lr scheduler (#28467)

vincentqb committed 4 years ago

d1d2358d

remove extraneous type information from torch.matrix_rank documentation (#28479)

ngoldbaum committed 4 years ago

9d767db4

Updating submodules

svcscm committed 4 years ago

e80e42cb

change empty range tolorrance logging

Xinyi Zhang committed 4 years ago

2f162842

Update Dockerfile

cerberusdevops committed 4 years ago

e9336b04

Update C++ torch::nn parity table for LayerNorm (#28484)

Will Feng committed 4 years ago

e28e38e8

Prepack folding for conv2d (#27119)

jerryzh168 committed 4 years ago

e280f93e

Migrate `smooth_l1_loss` from the TH to Aten (CPU & CUDA) (#27962)

andreaskoepf committed 4 years ago

be3808d3

Move complex extension test to c10 (#28208)

smessmer committed 4 years ago

ee920b92

Fix codegen for out operators (#28184)

smessmer committed 4 years ago

0f556b62

Allow migrating factory methods to c10 (#28183)

smessmer committed 4 years ago

b47d658d

Fix overload names (#28182)

smessmer committed 4 years ago

005d6ea4

Add unsupported types to schema type parser (#28181)

smessmer committed 4 years ago

a94bf1d3

Newer Older