strisunshinewentingwang/pytorch

Pull Requests Commits

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

pyi codegen - removing byte-for-byte-compatibility hacks (sorting overloads) (#49056)

bdhirsh committed 3 years ago

33a9b14d

pyi codegen - removing byte-for-byte compatibility hacks (#49055)

bdhirsh committed 3 years ago

b94ec8c9

pyi cleanup (#49054)

bdhirsh committed 3 years ago

9920adeb

Extra sampling of record function events [resend] (#49114)

Ilia Cherniavskii committed 3 years ago

db5e5b43

Fix structured kernel codegen (#49244)

smessmer committed 3 years ago

1cb5aa6c

[quant][graphmode][fx][fix] Fix typo in fusion (#49183)

jerryzh168 committed 3 years ago

2a3bb1ce

fix backwards compatibility for #48711 and its revert (#49240)

agolynski committed 3 years ago

796b2677

Expose run_async function on torch::jit::Method (#48607)

ShijunK committed 3 years ago

f965b0fc

Tuple Slice with both negative and positive stepped size (#48660)

tugsbayasgalan committed 3 years ago

42c78ed7

Improve new_group example in the context of SyncBatchNorm (#48897)

rohan-varma committed 3 years ago

c0a08450

[PyTorch Mobile] Record dtypes for tensors used in kernel function implementations (#48826)

dhruvbird committed 3 years ago

f10b53d9

Drop FutureNCCL in favor of vanilla CUDAFuture (#49014)

lw committed 3 years ago

f204f77e

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

generatedunixname89002005325676 committed 3 years ago

dcd1e3d7

Bring fast_nvcc.py to PyTorch OSS (#48934)

samestep committed 3 years ago

2bb2f641

add additional arm64 checker in cmake files (#48952)

Rong Rong committed 3 years ago

88b3d337

Revert D25428587: [pytorch][PR] add additional interpolation modes for torch.quantile

heitorschueroff committed 3 years ago

2f1d1eb7

Make CUDAFuture remember and restore current device in callback (#48789)

lw committed 3 years ago

5ab90b2f

Newer Older