strisunshinewentingwang/pytorch

Pull Requests Commits

Fix formatting error in `set_deterministic` documentation (#49136)

kurtamohler committed 3 years ago

fdadfb6e

[fx] Add constant folding pass (#48443)

jfix71 committed 3 years ago

38ed3985

Use group.WORLD appropriately in process group initialization. (#48767)

pritamdamania committed 3 years ago

f2ba3c16

Update pipeline API to accept arbitrary sequence of Tensors and not just Tuple (#48467)

pritamdamania committed 3 years ago

dc4db955

fix slow windows test (#49258)

agolynski committed 3 years ago

33b7970d

[pt] Replace size(dim) with sizes()[dim] (#49255)

Hao Lu committed 3 years ago

cd927875

Remove unused reconstruct_scopes function (#48822)

cccclai committed 3 years ago

717f31d9

[te] Use c10::ScalarType utility functions in te::Dtype (#49148)

bertmaher committed 3 years ago

dc92f25b

[te] Use Dtype::is_signed instead of an ad hoc local predicate. (#49147)

bertmaher committed 3 years ago

eaac2819

[te] Fix clamp with uint8 args (#49143)

bertmaher committed 3 years ago

ae88d25c

Fix "Missing return statement" mypy error (#49276)

malfet committed 3 years ago

8999915a

Revert D25434956: [JIT] Use `is_buffer` in `BufferPolicy::valid`

xw285cornell committed 3 years ago

b5b8fe98

[shape inference] fix ConstantFill

ChunliF committed 3 years ago

693e9086

[PyTorch] Remove native::zeros reference in TensorIndexing (#49117)

iseeyuan committed 3 years ago

8d58362f

Enable LayerNorm test cases

venkatacrc committed 3 years ago

635f1cd1

[JIT] Fix toIValue handling of AttributeError when casting ClassType (#49188)

James Reed committed 3 years ago

76d41c80

[Gradient Compression] Minor update of the comments on PowerSGD. (#49246)

Yi Wang committed 3 years ago

29f0fa36

Additional validation for DistributedSampler. (#48865)

pritamdamania committed 3 years ago

21c38e17

[te] Add BitCast to the IR (#49184)

bwasti committed 3 years ago

6b786446

Enabled Scalar lists (#48222)

Iurii Zdebskyi committed 3 years ago

5716b7db

inline `has` function for DispatchKeySet (#49191)

ngimel committed 3 years ago

bfce69d6

[FX] Move none assignments to same line (#49209)

James Reed committed 3 years ago

53aa9b8c

Add tensorpipe agent tests to multigpu tests. (#49210)

pritamdamania committed 3 years ago

2f359e7d

Modify Pipe to return an RRef. (#47829)

pritamdamania committed 3 years ago

df027bfd

[PyTorch] Fix getCustomClassType() perf (#48981)

swolchok committed 3 years ago

c6147ae4

Updated derivative rules for complex QR decomposition (#48489)

IvanYashchuk committed 3 years ago

6c1b405a

[PyTorch] avoid unnecessary call to empty_tensor_restride in empty() (#48211)

swolchok committed 3 years ago

e3542d2c

Reduce kineto logging (#49216)

Ilia Cherniavskii committed 3 years ago

4bc4ec26

Enable torch.where() to support Float16 & BFloat16 type inputs (#49004)

Kiyosora committed 3 years ago

15200e38

pyi codegen refactor - no need to group python signatures by overload name (#49057)

bdhirsh committed 3 years ago

218eaf4b

Newer Older