strisunshinewentingwang/pytorch

Pull Requests Commits

bypass `getDeviceFromPtr` check when device is known (#36714)

Emilio Castillo committed 4 years ago

f4183395

Have DeviceType available in torch namespace (#38036)

anjali411 committed 4 years ago

8e07b75c

Revert D21511048: [pytorch][PR] .circleci: Improve docker image build workflow

malfet committed 4 years ago

7c2853be

[ONNX] Fix pow op export (#38065)

neginraoof committed 4 years ago

333e29c4

fix sample code (#38002)

srayuws committed 4 years ago

19d6e32e

.circleci: Improve docker image build workflow (#37976)

seemethere committed 4 years ago

def9f15b

test_linspace : remove explicit for-loop (#38191)

kshitij12345 committed 4 years ago

a37b8651

Pin flake8 to 3.7.9 (#38269)

Shen Li committed 4 years ago

c6b28440

[JIT] Expose magic methods on script::Object (#38167)

James Reed committed 4 years ago

a553935e

[JIT] Disallow plain List type annotation without arg (#38130)

Meghan Lele committed 4 years ago

1456515f

Using LoadLibraryEx and LOAD_LIBRARY_SEARCH_* flag for loading DLLs o… (#37763)

peterjc123 committed 4 years ago

00f3790a

Add instance methods tensor.isnan(), tensor.isinf(), tensor.isfinite() (#37942)

ShawnZhong committed 4 years ago

5f9b9036

Lazily initialise thread local num_threads value (#37461)

peterbell10 committed 4 years ago

5137827a

[pyfi] override TP2 networkx -> PyFI networkx (#37764)

Tao Wu committed 4 years ago

08c3339e

DOC: add BFloat16 dtype and BFloat16Tensor (#37051)

mattip committed 4 years ago

c3191367

Migrate CPU tril, triu, masked_fill to c10::complex (#37897)

zasdfgbnm committed 4 years ago

b290da0e

If we're building on C++17, use actual "if constexpr" (#38154)

smessmer committed 4 years ago

77d8a448

Inverse logic of persistent set and prevent use in jit (#38131)

albanD committed 4 years ago

3569c596

Remove codegen for IntArrayRefStride, which isn't used. (#38072)

gchanan committed 4 years ago

f314d9a0

Macro generate ScalarTypeToCPPType, including all ScalarTypes. (#38071)

gchanan committed 4 years ago

fe53b525

Kill resize-ing and zero-ing from codegen. (#37958)

gchanan committed 4 years ago

c26dde96

add missing include file for fake_nnpi_ops_utils (#38215)

Hector Yuen committed 4 years ago

ebad4e46

Delete torch/__init__.pyi, deferring to direct extension stubs (#38157)

ezyang committed 4 years ago

6edf3403

Add per-device allocator object in CUDACachingAllocator (#37567)

mtbrandy committed 4 years ago

6f396e18

add dtype checking for gather and scatter (#38025)

cloudhan committed 4 years ago

324dc162

fixing build failures with USE_NATIVE_ARCH ON (#35359)

Protonu Basu committed 4 years ago

503be4e0

explain redundant branch/tag filters (#38169)

kostmo committed 4 years ago

f3e620ee

[Resubmit] Migrate AT_DISPATCH_ALL_TYPES_AND_COMPLEX_AND3 to c10::complex (#38144)

zasdfgbnm committed 4 years ago

5077518c

remove internal file logging.h (#38182)

Hector Yuen committed 4 years ago

26928b16

[TensorExpr] remove Let and LetStmt in favour of binding in Block (#37606)

nickgg committed 4 years ago

33f4fca1

Newer Older