strisunshinewentingwang/pytorch

Pull Requests Commits

fix dropout AD & rename range to rangelist (#17691)

Ailing Zhang committed 5 years ago

fefaebab

enable use of MIOpen for depthwise convolutions (#17685)

lcskrishna committed 5 years ago

36e0d39f

Implement a Caffe2 standalone LSTM operator (#17461)

ahhegazy committed 5 years ago

bfe7a58f

Fix nll_loss crash on cpu where ignore_index is out of bounds (#17328)

soumith committed 5 years ago

a478d416

Add '--hip-clang-launch' to favor <<<>>>-based launch. (#17686)

iotamudelta committed 5 years ago

288e1fbd

Improve caching allocator for Pascal and newer GPUs. (#17120)

colesbury committed 5 years ago

079093a6

Turn the Half::from_bits into a constexpr function to avoid unresolve… (#17661)

dlibenzi committed 5 years ago

8420a202

Remove Expect Files from python / tracing / script interop

Elias Ellison committed 5 years ago

7fc3aa8c

Enable apex on Windows

peterjc123 committed 5 years ago

b219882c

bump docker build to upgrade magma to 2.5.0 (#17674)

soumith committed 5 years ago

f176450d

refactor caffe2 operator constructors - 1/9 (#17082)

smessmer committed 5 years ago

54c4b5a4

Expose cuda kernel for caffe2::GenerateProposals

smessmer committed 5 years ago

910519e4

print warnings when DNNLOWP_16 or DNNLOWP_ROWWISE_16 engine is used (#17176)

jspark1105 committed 5 years ago

aea8dd83

Fix XOutput/XOutputTensor for ivalue based c2 operators (#17599)

smessmer committed 5 years ago

8569d9cb

Fix InputSize/OutputSize for ivalue based operators (#17579)

smessmer committed 5 years ago

c7db0b35

Fix clamp fusion on missing limits (#17533)

wanchaol committed 5 years ago

173561ff

int32 indexing for Tensor Iterator Reduction (#17428)

jjsjann123 committed 5 years ago

a87eeec9

Removed all usages of TH_Index_Base (#17591)

izdeby committed 5 years ago

32576082

PyTorch/Caffe2 tensor interop in Python (#17190)

Dmytro Dzhulgakov committed 5 years ago

dec116e9

Fixed typo in aten/src/ATen/native_parse.py (#17641)

wkcn committed 5 years ago

244d3309

Remove GPU dependency from ProfileObserver (#17592)

Martin Schatz committed 5 years ago

5b835682

Don't make factory methods create a tensor and then immediately copy it (#17565)

Brennan Vincent committed 5 years ago

6a297b86

Fixed typo in torch/functional.py w/r/t broadcast_tensors (#17642)

jrichterpowell committed 5 years ago

7a51c03a

Change fake tqdm constructor to match real tqdm (#17636)

bryanhe committed 5 years ago

01977c0a

Mark native_functions as matched if uncaptured by JIT (#17631)

cpuhrsch committed 5 years ago

ef7ddcd2

Ban std::array from native_functions.yaml

cpuhrsch committed 5 years ago

80927fc0

Remove more usages of BoolTensor and IndexTensor from native_functions.yaml

cpuhrsch committed 5 years ago

416474a7

Implement kthvalue in ATen (#17544)

t-vi committed 5 years ago

c6715eda

Change vml.h to support sizes greater than 2**32 - 1

cpuhrsch committed 5 years ago

43f94077

msvc_fixes (#17201)

ArutyunovG committed 5 years ago

2336f0ba

Newer Older