strisunshinewentingwang/pytorch

Pull Requests Commits

SparseCsr: Fix some uses of deprecated Tensor methods (#58990)

peterbell10 committed 3 years ago

aaccdc39

OpInfo: tensor_split: port remaining method_test entries (#59133)

kshitij12345 committed 3 years ago

6ee9466d

Replace `dim_apply` with `TensorIterator` (#58656)

peterbell10 committed 3 years ago

96c549d1

OpInfo: renorm (#59079)

kshitij12345 committed 3 years ago

cab65ea3

[special] Add `i1` and `i1e` (#56352)

kshitij12345 committed 3 years ago

5c189946

[TensorExpr] Add NNC lowerings for `aten::view`, `aten::reshape` and `aten::expand_as`. (#59157)

Mikhail Zolotukhin committed 3 years ago

27009d61

make vector_norm backward call norm_backward (#59135)

ngimel committed 3 years ago

355b2443

OpInfo: tril, triu (#59145)

kshitij12345 committed 3 years ago

9fc0c5a5

avoid explicitly casting low precision inputs to fp32 in norm (#59134)

ngimel committed 3 years ago

1871d4e6

OpInfo: fill_ (#59138)

kshitij12345 committed 3 years ago

d68df542

[NNC] Added triangular_solve external call + fixed permute (#59131)

Chillee committed 3 years ago

a4278203

OpInfo: where (#58349)

kshitij12345 committed 3 years ago

c9af4c26

[c10d] Split custom class bindings out of python binding code (#58992)

suo committed 3 years ago

b977a3b6

[iOS GPU] Add debug information to track memory allocation exception (#59112)

xta0 committed 3 years ago

ab372ba5

CUDA support in the CSR layout: sparse_to_dense/add_sparse_csr (#59011)

aocsa committed 3 years ago

41054f2a

Use some c10::ThreadLocal to avoid crashes on old Android toolchains (#59017)

dreiss committed 3 years ago

9c83e416

Include Macros.h in ThreadLocal

dreiss committed 3 years ago

4b3d17c0

OpInfo: `fmod` and `remainder` (#57941)

krshrimali committed 3 years ago

0c1420aa

Revert D28700259: [pytorch][PR] Migrate nonzero from TH to ATen (CPU)

ngimel committed 3 years ago

657b75d1

Move remaining \*Sort\* in `THC` to `ATen` (#58953)

eqy committed 3 years ago

4e543d01

Add peephole for len(x.size()) (#59051)

eellison committed 3 years ago

f3aa61b9

Add more shape functions for mobilenet (#58932)

eellison committed 3 years ago

b9dc5186

Switch symbolic shape registry to operator map' (#58890)

eellison committed 3 years ago

0ebc6653

[JIT] Disable Complete Shape Inlining For Testing Purposes (#56966)

eellison committed 3 years ago

d8cbba3e

Add unary/binary ops necessary for mobilenet (#56828)

eellison committed 3 years ago

f66fbb1e

Use dataclasses to simplify ShardingSpec (#58893)

pritamdamania committed 3 years ago

40f851c5

[quant][refactor tests] Move qtensor serialization tests from test_deprecated_jit (#59089)

supriyar committed 3 years ago

89d78851

[quant][refactor tests] Clean up test_quantization.py (#59088)

supriyar committed 3 years ago

886a2ddc

TensorIteratorReduce: Avoid tensor operations in parallel_for (#58655)

peterbell10 committed 3 years ago

f993ceff

Back out "[pytorch][PR] ENH Adds dtype to nn.functional.one_hot" (#59080)

jbschlosser committed 3 years ago

ef32a29c

Newer Older