strisunshinewentingwang/pytorch

Pull Requests Commits

Updated codegen to remove TensorOptions from generated code

izdeby committed 5 years ago

02370f11

Prepare templates

izdeby committed 5 years ago

38008fbc

Updated DispatchKeyExtractor to expect TensorOptions

izdeby committed 5 years ago

ddd03349

Use default scale/zero_point in fake_quantize module instead of None (#32318)

jerryzh168 committed 5 years ago

8c1268aa

anjali411 committed 5 years ago

5b815d98

Don't dispatch to integral types in smooth_l1_kernel

zasdfgbnm committed 5 years ago

78d8f691

use gtest asserts in ProcessGroupGlooTest instead of other checks (#32138)

rohan-varma committed 5 years ago

6a5a55d5

cap the maximum depth of bailout chains at 1 (#32073)

Krovatkin committed 5 years ago

4968bc24

Updating submodules

svcscm committed 5 years ago

61a2b341

fix testSend and testRecv in ProcessGroupGlooTest (#32134)

rohan-varma committed 5 years ago

904ab092

add lock for ncclCommAbort (#31901)

zhaojuanmao committed 5 years ago

7a9c920b

fix spelling mistake: excpected -> expected

your-average-bear committed 5 years ago

91bdb872

Register RoIAlignRotated with C10

SuperIRabbit committed 5 years ago

ef5ae482

remove unused code after refactoring optimizations into profiling-sensitive and profiling-insensitive (#32106)

Krovatkin committed 5 years ago

b79030d6

Enhancing the test (#32321)

Xintao Chen committed 5 years ago

c2761490

classic fixed-point liveness

Krovatkin committed 5 years ago

53708e21

Add CI scripts for Custom Build (#32316)

xta0 committed 5 years ago

8c8bd79f

Eliminate exception throwing code from dispatch call sites (#32168)

ezyang committed 5 years ago

34c751c2

Out-of-line construction of OperatorName. (#32121)

ezyang committed 5 years ago

b85dbe8f

Move error reporting code out-of-line from header. (#32118)

ezyang committed 5 years ago

36d09197

Make an assert on a hotpath trigger only in DEBUG mode. (#32117)

ezyang committed 5 years ago

7b739077

Fix weight backward for cudnn conv of large tensor (#31889)

zasdfgbnm committed 5 years ago

8746f90c

For ppc64le, stop presenting the python 2.7 builds (we will no longer… (#32315)

dncliss committed 5 years ago

b26ee541

Pin Pillow to latest and use a torchvision that works with it (#32290)

hugovk committed 5 years ago

cd99b370

Logical condition reduction (#32201)

gaurav1086 committed 5 years ago

f94aab45

out variant for native_batch_norm forward (#29192)

vadimkantorov committed 5 years ago

14548c2d

reimplement __torch_function__ overrides for torch.functional using inline logic (#32194)

ngoldbaum committed 5 years ago

bab87e4b

Creating callUnboxedWithDispatchKey method (#32198)

Xintao Chen committed 5 years ago

7df5dc27

Support shape inference and lowering of SparseLengthsWeightedSumFused4BitRowwise (#32257)

Yinghai Lu committed 5 years ago

d75b6b3f

Updating submodules

svcscm committed 5 years ago

f3b62d4b

Older