strisunshinewentingwang/pytorch

Pull Requests Commits

[nnc] llvm_codegen quantization types for vectype (#68736)

IvanKobzarev committed 2 years ago

7d8a79b6

Sparse CSR CUDA: Add block torch.addmm when mat1 is sparse (#68707)

IvanYashchuk committed 2 years ago

b28ddd72

Cleanup old TD logic (#68842)

malfet committed 2 years ago

b5b62b34

[SR] Use std::vector::reserve for StaticModule constants (#68834)

donaldong committed 2 years ago

d9f3feb5

Update Documentation to Make CUDA Call Explicit (#67973)

gauenk committed 2 years ago

8fb9ce49

[Quant] Refactor handling of FixedQParams operators (#68143)

andrewor committed 2 years ago

79b67d9a

All get_attr node to be in64 type (#68818)

Shirong Wu committed 2 years ago

998daf44

[BE] Simplify magma installation logic (#68778)

malfet committed 2 years ago

78dce417

Fix `test_svd_errors_and_warnings` warning message when cuda >= 11.5 (#68683)

xwang233 committed 2 years ago

2cd48d14

Revert D32611368: [pytorch][PR] Initial version of general convolution_backward

osalpekar committed 2 years ago

8e343ba5

Add API usage logging to ShardedTensor and fix a few tests. (#68771)

pritamdamania committed 2 years ago

84047ff3

Populate operator_input_sizes_ (#68542)

qihqi committed 2 years ago

959cb031

[pytorch] Fix loading from checkpoint after "maximize" flag was introduced in SGD (#68733)

asanakoy committed 2 years ago

c0e6dc9a

.circleci: Remove migrated CUDA 10.2 build (#68782)

seemethere committed 2 years ago

73f494d6

Making norms inputs independent (#68526)

Samantha Andow committed 2 years ago

23288fda

Require CMake 3.13 when building with Ninja (#68731)

peterbell10 committed 2 years ago

e7e1b761

Added additional string to search cpu flags for vnni detection (#67686)

Caspar van Leeuwen committed 2 years ago

3282386a

[dist_quant] change op registration to each file instead (#68797)

wanchaol committed 2 years ago

98e51895

Initial version of general convolution_backward (#65219)

jbschlosser committed 2 years ago

445b31ab

[quant][graphmode][fx] Add support for specifying reference quantized module mapping in backend_config_dict (#68227)

jerryzh168 committed 2 years ago

a31aea8e

[sparsity] Fix for the failing pruner test (#68794)

z-a-f committed 2 years ago

b845b987

[PyTorch][3/N] Enable the rest forward spec options for ShardedEmbedding and ShardedEmbeddingBag. (#67799)

fduwjj committed 2 years ago

d6a68e0b

Add OpInfos for parcel Activation Functions I (#68521)

saketh-are committed 2 years ago

5d300e76

fix typos in jit_language_reference.rst (#68706)

Yutaro-Sanada committed 2 years ago

74e6d2ce

[sparsity] Fix GPU training for sparsity (#66412)

z-a-f committed 2 years ago

e7d8f096

Fix strict aliasing rule violation in bitwise_binary_op (#66194)

pbialecki committed 2 years ago

0b067412

[sparsity] Fix and enable the pruning tests (#66411)

z-a-f committed 2 years ago

d176c82b

Add linalg.solve_triangular (#63568)

lezcano committed 2 years ago

b46c89d9

refactor: update f-string for swa.utils.py (#68718)

nhankiet committed 2 years ago

a2e35e16

[Dist CI][BE] c10d gloo tests run in subprocess (#68504)

rohan-varma committed 2 years ago

9554ebe4

Older