strisunshinewentingwang/pytorch

Pull Requests Commits

Fixing MAGMA with HIP issues (#56448)

amathews-amd committed 3 years ago

ed2104fe

Local lint fixes - missing steps, pin to bash (#56752)

davidriazati@fb.com committed 3 years ago

0424f6af

Fix bug in emitUse to drop all values that are marked as drop (#56652)

tugsbayasgalan committed 3 years ago

6de1d9b2

[PyTorch][Edge] Add v4 and v5 models and remove unused model (#56751)

cccclai committed 3 years ago

2e4c68a7

Add a new API replace_input_with to node.py (#55887)

chenccfb committed 3 years ago

798dd466

Fix instance norm input size validation + test (#56659)

jbschlosser committed 3 years ago

7d2a9f2d

[Pytorch Edge] Prepack folding for functions besides forward (#56081)

JacobSzwejbka committed 3 years ago

7e9f7fb9

[c10d] Increment sequence numbers on collectives. (#55718)

rohan-varma committed 3 years ago

7ff1990c

Revert D27902824: static runtime support for fb::equally_split

xw285cornell committed 3 years ago

ed0a0c35

To add single and chained learning schedulers to docs (#56705)

iramazanli committed 3 years ago

d1fe68e7

Use JIT Plug-in for coverage to cover JIT'd functions and methods (#56310)

janeyx99 committed 3 years ago

88bd0510

Make sure full backward hook fire when no input requires grad (#56693)

albanD committed 3 years ago

22b151a3

Add more RRef CUDA RPC tests (#56757)

mrshenli committed 3 years ago

acca89e2

Added support for uppercase letters in torch.einsum (#56475)

heitorschueroff committed 3 years ago

369e8bc4

Add CUDA support to a user-created torch.futures.Future (#56517)

lw committed 3 years ago

15ca379b

Allow to specify a set of device for CUDAFuture (#56515)

lw committed 3 years ago

58d12eb7

add hardtanh(0,6) to the set of MKLDNN fusible ops for mobilenetv2 (#56203)

Krovatkin committed 3 years ago

d6a25a58

[PyTorch] Migrate hacky wrapper removal to borrow_from_optional_tensor (#56648)

swolchok committed 3 years ago

7b7a4750

[PyTorch] Add & document borrow_from_optional_tensor (#56647)

swolchok committed 3 years ago

f2fd91cc

addmm CPU inplace implementation shouldn't resize an input tensor (#56452)

imaginary-person committed 3 years ago

02c3e6d9

Fix: Compare input against beta * threshold in softplus backwards (#56484)

jbschlosser committed 3 years ago

e5fda07e

Some simple optimizations (#51831)

cyyever committed 3 years ago

83c23703

Torchelastic: make process failure init error non-fatal (#56739)

aivanou committed 3 years ago

0a72904a

static runtime support for fb::equally_split (#56565)

Edvard Ghazaryan committed 3 years ago

a4e47ea1

moved more lowerings over (#55372)

Chillee committed 3 years ago

7c50852a

Consolidate nondeterministic error tests (#55631)

kurtamohler committed 3 years ago

1f04494c

[torch.package] is_from_package check (#56729)

Lilyjjo committed 3 years ago

88deea4e

Revert "Revert [ONNX] Redesign inplace conversion" (#56675)

BowenBao committed 3 years ago

913f1f75

CPU Convolution benchmark harness for some popular models (#56455)

bertmaher committed 3 years ago

461e887d

Move windows testers to previous image (#56626)

malfet committed 3 years ago

f84a5010

Newer Older