strisunshinewentingwang/pytorch

Pull Requests Commits

Propagate information on torch_shm_manager execl failure to parent process (#57310)

vec4f committed 3 years ago

e68c46bb

Address temp file/bind race condition in torch_shm_manager (#57309)

vec4f committed 3 years ago

2c2aa9e0

Make c10::TempFile non-copyable but movable (#57308)

vec4f committed 3 years ago

7eed5410

Propagate information on torch_shm_manager failures to parent process (#57307)

vec4f committed 3 years ago

788aefd7

static graph api skeleton (#54995)

zhaojuanmao committed 3 years ago

3f819128

refactor autograd_hook (#54981)

zhaojuanmao committed 3 years ago

5f2b9b1d

Add pybind interface to caffe2 quantization server (#57330)

Adolfo Victoria committed 3 years ago

81ef82e5

Static Runtime support for aten::matmul (#57291)

Edvard Ghazaryan committed 3 years ago

e62cdae4

Update DLPack to 0.4 (#55365)

Emilio Castillo committed 3 years ago

0a9c9cc6

[PyTorch][jit] Don't allow create() on singleton types (#56807)

swolchok committed 3 years ago

b87d3fa4

[fx splitter] Fix fusion group utility (#57280)

842974287 committed 3 years ago

d896d1f4

[nnc] Enable all fuser tests for cpu (#57332)

bertmaher committed 3 years ago

7c8a7efe

reduce inline autodiff threshold so we can caputre smaller fusions (#57062)

Krovatkin committed 3 years ago

d50a969f

[NNC] Make flatten transform in-place (#56629)

navahgar committed 3 years ago

e795f88d

Fix string_view::equals_ compilation by CUDA-11.3 (#57322)

malfet committed 3 years ago

b49e079a

[PyTorch] Include hip_runtime.h in macros.h (#57070)

swolchok committed 3 years ago

52805a0f

[JIT] Disable conv-add-relu fusion for cuDNN7 when model uses fp16 (#56579)

desertfire committed 3 years ago

c9714016

refactor autocast to be extensible for devices (#57104)

leslie-fang-intel committed 3 years ago

731cc472

Add supported backward_dtype to OpInfo (#56156)

Rong Rong (AI Infra) committed 3 years ago

095c328d

Revert D27582224: [pytorch][PR] Automated submodule update: FBGEMM

malfet committed 3 years ago

e08303c7

Revert D28114231: [pytorch][PR] Automated submodule update: FBGEMM

malfet committed 3 years ago

0dddfbf3

Remove unused forward AD flag (#57058)

albanD committed 3 years ago

95dc2b6e

Improve perf for forward AD view handling (#57057)

albanD committed 3 years ago

83f18671

fix InplaceOrView implementation for manual functions (#57152)

albanD committed 3 years ago

b016bc1c

Fix use of allow_tensor_metadata in view variable creation (#57069)

albanD committed 3 years ago

c91bd25e

make external codegen aware of autogen'd composite kernels (#56960)

bdhirsh committed 3 years ago

6fa1d880

[special] Add `xlog1py` (#55138)

kshitij12345 committed 3 years ago

d4ddb477

Expand Kineto profiler support (part 1) (#57333)

Ilia Cherniavskii committed 3 years ago

5b3e7638

quote str kwarg values in `test_ops.py::TestCommon::test_jit_alias_remapping` (#57120)

pmeier committed 3 years ago

db32b695

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

CodemodService FBSourceClangFormatLinterBot committed 3 years ago

df69b0d0

Newer Older