strisunshinewentingwang/pytorch

Pull Requests Commits

Automated submodule update: FBGEMM (#57342)

facebook-github-bot committed 3 years ago

264db195

Automated submodule update: FBGEMM (#55347)

facebook-github-bot committed 3 years ago

54469e15

Static runtime support for fb::expand_dims (#57282)

Edvard Ghazaryan committed 3 years ago

b3e18024

[torch/deploy] opt torch/csrc/depoy into autofromatting

suo committed 3 years ago

e31b67f5

Fix a numerical issue of CUDA channels-last SyncBatchNorm (#57077)

xwang233 committed 3 years ago

ac72881f

Ignore more compiler warnings, unify WERROR options (#56630)

davidriazati@fb.com committed 3 years ago

c44cbc63

Revert "Remove sync for randperm on small tensors. (#54113)" (#57299)

ngimel committed 3 years ago

65968ab8

[kineto] Deprecate ClientTraceActivity and merge it with GenericTraceActivity (#56743)

chaekit committed 3 years ago

49dbe179

port neg to structure kernel (#57212)

Rong Rong (AI Infra) committed 3 years ago

16fc18bf

Fix sort for slow gradcheck (#57192)

soulitzer committed 3 years ago

99516120

[torch] Add backward support for segment reduce (CPU only)

Serhat Yilmaz committed 3 years ago

e27740b3

[torch/debuggability] use log.info() in addition to print() in timeoutguard (#57296)

Aapo Kyrola committed 3 years ago

d1def931

[RPC Framework] Expose a Python API for device map getter (#57179)

Yi Wang committed 3 years ago

c2fbd967

[package] PackageExporter `__import__` logic to not parse dynamic cases (#57283)

Lilyjjo committed 3 years ago

2c6f5e8a

Added OpInfos for sub & mul (#56227)

imaginary-person committed 3 years ago

6ed90ed1

Update compare_set docs (#57203)

H-Huang committed 3 years ago

149000c3

Fix path handling on Win32 in rendezvous.py (#57000)

skyline75489 committed 3 years ago

e31265df

[fx minimizer] Add an option to minimizer to allow return all intermediate results (#57279)

842974287 committed 3 years ago

a6fa6a6c

Add compare_set to trampoline class, add typing and formatting (#57191)

H-Huang committed 3 years ago

95f393f2

[torch/deploy] Minor housekeeping in interpreter_impl

suo committed 3 years ago

be0ca00c

Remove distutils (#57040)

davidriazati@fb.com committed 3 years ago

4b96fc06

Add torch_cpu specific flag for debug info (#57190)

davidriazati@fb.com committed 3 years ago

21be40b3

[PyTorch] Allocate correctly-sized output tensor in addmm_cuda (#56033)

swolchok committed 3 years ago

d3ffe9ab

[PyTorch] Reduce move overhead in inferExpandGeometry (#56032)

swolchok committed 3 years ago

dd9f4c8c

[PyTorch] Migrate copy_ to borrow input/output (#56031)

swolchok committed 3 years ago

fb2f3cd1

[PyTorch] Make as_strided_ use_const_ref_for_mutable_tensors (#55875)

swolchok committed 3 years ago

a1d2bd56

fix: index_fill_ formula to support duplicate indices (#57101)

kshitij12345 committed 3 years ago

ac86e0a0

Fix for derivative of sinc(x) when x is positive but very very small (#56986)

kbrose committed 3 years ago

ec86f96e

[Distributed test]Enable ddp_control_flow tests for ROCm (#57159)

jaglinux committed 3 years ago

fd67088a

Support type inference of nn.Module methods using PDT (#57165)

nikithamalgifb committed 3 years ago

2e2c0099

Newer Older