strisunshinewentingwang/pytorch

Pull Requests Commits

Support AMP in nn.parallel (#43102)

anoidgit committed 4 years ago

f02753fa

[serialize] Expose zip file alignment calculation functions (#43531)

Dmytro Dzhulgakov committed 4 years ago

cbdaa20c

force pytorch tensors to contiguous before calling c2 ops

ngimel committed 4 years ago

d1d32003

Fix "save binary size" steps (#43529)

malfet committed 4 years ago

675f3f04

Properly format db.h and db.cc (#43027)

Sean Lynch committed 4 years ago

f80b695a

[quant][graphmode[fx][test][refactor] Refactor tests for graph mode quantization on fx (#43445)

jerryzh168 committed 4 years ago

7b243a4d

[pytorch] add option to include autograd for code analyzer (#43155)

ann-ss committed 4 years ago

87905b58

[quant] Support set API for EmbeddingBag quantization (#43433)

supriyar committed 4 years ago

284ff047

[2/3][lite interpreter] add metadata when saving and loading models for mobile

husthyc committed 4 years ago

e37f871e

Update quantize_jit to handle new upsample overloads (#43407)

dreiss committed 4 years ago

ed8b08a3

Reland of benchmark code (#43428)

albanD committed 4 years ago

e08e93f9

[ROCm] allow .jenkins/pytorch/test.sh to run on centos (#42197)

jeffdaily committed 4 years ago

4cfac340

Implement JIT Enum type serialization and deserialization (#43460)

gmagogsfm committed 4 years ago

35a36c12

Enable torch.cuda.memory typechecking (#43444)

malfet committed 4 years ago

0fa99d50

[quant] Add benchmarks for quantized embeddingbag module (#43296)

supriyar committed 4 years ago

7024ce8a

Add lite SequentialSampler to torch mobile (#43299)

ann-ss committed 4 years ago

7cc1efec

Implement batching rules for basic arithmetic ops (#43362)

zou3519 committed 4 years ago

c972e623

Enable torch.cuda.nvtx typechecking (#43443)

malfet committed 4 years ago

db78c07c

[Codemod][FBSourceClangFormatLinter] Daily `arc lint --take CLANGFORMAT`

generatedunixname89002005287564@sandcastle105.ftw3.facebook.com committed 4 years ago

2f9c9796

Add alias torch.negative to torch.neg. (#43400)

hameerabbasi committed 4 years ago

c4e84165

[fx] add type annotations (#43083)

zdevito committed 4 years ago

1f0cfbaa

[fx] enabling typechecking of fx files (#43082)

zdevito committed 4 years ago

b349f58c

remove prim::profile and special-casing (#43160)

Krovatkin committed 4 years ago

a97ca93c

[DPER3] Separate user embeddings and ad embeddings in blob reorder

ChunliF committed 4 years ago

d70b263e

Creates test_tensor_creation_ops.py test suite (#43104)

Mike Ruberry committed 4 years ago

4dc8f3be

Fix ToC Link (#43427)

scopatz committed 4 years ago

35351ff4

Fix bugs in vec256_float_neon.h (#43321)

AWSNB committed 4 years ago

e4af45f3

Enable input pointer caching in XNNPACK integration. (#42840)

kimishpatel committed 4 years ago

b003f2cc

Change quantizer to account for input tensor's memory format. (#42178)

kimishpatel committed 4 years ago

b52e6d00

Reduce number of `prim::profile` (#43147)

Krovatkin committed 4 years ago

b1d31428

Older