strisunshinewentingwang/pytorch

Pull Requests Commits

[JIT] Add CUDNN Conv-Add-Relu fusion for Frozen Model Optimization (#52102)

desertfire committed 3 years ago

4626886f

cmake: auto-detect ccache to speed up developer builds (#49389)

leezu committed 3 years ago

90bbe0b3

Test tools/test_history.py (#54259)

samestep committed 3 years ago

a95abc46

Use shard file if present, improve functions used for sharding (#54210)

janeyx99 committed 3 years ago

0645e2b4

Remove usage of onEachDevice from legacy profiler (#54125)

Ilia Cherniavskii committed 3 years ago

3b1e3103

Cleanup mypy lint job (#54260)

malfet committed 3 years ago

d85faf8d

Fixed the size of the workspace array in functions calling MAGMA (#54009)

IvanYashchuk committed 3 years ago

04a25060

Final kernel launch checks (#54214)

r-barnes committed 3 years ago

f0056f89

cleanup static_cast of AutogradMeta (#54103)

albanD committed 3 years ago

cc92117a

properly make AutogradMeta/DifferentiableViewMeta attributes internal (#54102)

albanD committed 3 years ago

004db373

Remove legacy from optional-related function names (#54101)

albanD committed 3 years ago

09b4af2f

Add size check for forward grads (#54100)

albanD committed 3 years ago

a425eb21

make internal forwardAD methods on at::Tensor internal (#54099)

albanD committed 3 years ago

cba8516b

Add MyPY to lint GHA workflow (#54067)

malfet committed 3 years ago

a52e295c

[NNC] Adding API to distribute loops (#53865)

navahgar committed 3 years ago

4b2abc4b

[PyTorch] Rename XPLAT_MOBILE_BUILD to TEMPLATE_SELECTIVE_BUILD (#54217)

cccclai committed 3 years ago

dc358488

Resubmit: Adding parallel support for the LLVM backend. (#54122)

zheng-xq committed 3 years ago

9f86b656

Optimize alias_analysis node lookup (#54115)

tugsbayasgalan committed 3 years ago

444552e7

Add torch.linalg.vector_norm function (#51099)

kurtamohler committed 3 years ago

382a47b4

Added autograd support for torch.orgqr (#52637)

IvanYashchuk committed 3 years ago

564456ac

Add cusolver potrf and potrfBatched to the backend of torch.cholesky decomposition (#53104)

xwang233 committed 3 years ago

2f3b194d

Revert D27001339: Use type-erased union for Buffer.

Mike Ruberry committed 3 years ago

8caa7889

Use type-erased union for Buffer. (#322)

Lucas Hosseini committed 3 years ago

c618dc13

[distributed] add processgroup options as argument (#53663)

wanchaol committed 3 years ago

133000fe

[FX] Normalize torch. namespace ops (#53832)

James Reed committed 3 years ago

2d8795c5

Remove __get__ from Tensor stub. (#54208)

ezyang committed 3 years ago

72c7983f

[FX] Experimental type annotation pass using Python signatures (#53831)

James Reed committed 3 years ago

a27f46bb

[WIP] Function to retrieve inspect.Signature instances for PyTorch ops (#53830)

James Reed committed 3 years ago

255b103c

Revert D26907093: Add repeats to Timer.collect_callgrind(...)

pbelevich committed 3 years ago

0dc5abfa

[StaticRuntime] Fuse SigridTransforms + ListUnpack (#53920)

Hao Lu committed 3 years ago

ca429fed

Newer Older