strisunshinewentingwang/pytorch

Pull Requests Commits

Linter to avoid csv merge conflicts (#111163)

PaliC committed 1 year ago

af05fbb8

[AOTInductor] Improve validation for C++ wrapper codegen (#111102)

khabinov committed 1 year ago

8209bbbd

[logging] log exceptions when provided (#111164)

davidberard98 committed 1 year ago

898482f1

Public API for constructing NT with jagged layout from tensor list (#111078)

jcaip committed 1 year ago

4c016860

[PyTorch] AOTI: add CPU fast path in aoti_torch_empty_strided (#110877)

swolchok committed 1 year ago

a2c17a2b

[PyTorch] -DNDEBUG in inductor codecache builds (#110876)

swolchok committed 1 year ago

b85f8482

[export] Reland "Fix graph signature data model to list of specs." (#111136)

zhxchen17 committed 1 year ago

168bad5f

Quant: add weight int4pack mm kernel (#110914)

yanboliang committed 1 year ago

9980876c

add Half support for bernoulli on CPU (#104176)

CaoE committed 1 year ago

8713a1a3

Update sdp_utils functions to accept const& params (#111144)

drisspg committed 1 year ago

74b1f4f7

[Vulkan] Add the 2D case to Layernorm operator (#110796)

copyrightly committed 1 year ago

21dc1d25

[inductor]: Better debugging of `can_fuse` decisions with `TORCH_LOGS=fusion` (#110415)

jon-chuang committed 1 year ago

9c7f464e

[docs] export full aten opset (#111161)

avikchaudhuri committed 1 year ago

1208a447

define public API for torch.nn.utils (#111026)

ringohoffman committed 1 year ago

ad447283

Expand NT subclass to support SAM (#109123)

jbschlosser committed 1 year ago

8f90be44

[DCP] Remove _shard_tensor() call in load_sharded_optimizer_state_dict in optimizer.py (#111096)

wz337 committed 1 year ago

e0eaa95e

[aot_inductor] add a test with AOTInductor + TorchScript (#111124)

chenyang78 committed 1 year ago

6748a14a

Fix typo in mixed dtypes linear operator implementation. (#111127)

alexsamardzic committed 1 year ago

397deaa8

Revert "[inductor] Implement Fx graph caching to improve warm compilation time. (#103453)"

pytorchmergebot committed 1 year ago

7fbfa4e0

Add pypi required metadata to all wheels except linux (#111042)

atalman committed 1 year ago

f9053877

Disable cutlass_template on ROCm (#111132)

jataylo committed 1 year ago

94c9dbff

Reland #2 "[C10] PG observability hooks. (#108815, #110907)" (#111072)

wconstab committed 1 year ago

bb1424d4

[BE] Enable Ruff's Flake8 PYI018 (#111101)

isdanni committed 1 year ago

dede1e96

Added decorator `skipRocmIfTorchInductor` and skipped failing tests (#107760)

pragupta committed 1 year ago

53a9ac53

[Inductor] support channel last for xpu conv in inductor layout opt path (#111018)

guangyey committed 1 year ago

918054f4

fix: do not reshard parameters twice (#110948)

0x804d8000 committed 1 year ago

5ace9122

Move cuda driver exit handling from helpers to threads (#111061)

wconstab committed 1 year ago

aec0f98e

[BE] Enable Ruff's Flake8 PYI030 (#111103)

isdanni committed 1 year ago

2f53085f

Move at::{Refcounted,}MapAllocator to c10 (#109881)

peterbell10 committed 1 year ago

68a1219f

Revert "[export] Fix graph signature data model to list of specs. (#111017)"

pytorchmergebot committed 1 year ago

42b89aea

Newer Older