strisunshinewentingwang/pytorch

Pull Requests Commits

try dynamo in parallel

clee2000 committed 2 years ago

f14e8e13

Revert "match sdpa patterns from HF (#100609)"

pytorchmergebot committed 2 years ago

0585944e

[pt2] add meta for `linalg_lu_factor_ex` (#101375)

nkaretnikov committed 2 years ago

42e65a25

[GHF] Ignore flaky classification for pin updates (#101587)

malfet committed 2 years ago

cb734123

Test Reordering: Run previously failing tests first (#101123)

ZainRizvi committed 2 years ago

b1474019

use Bazelisk to fetch Bazel in CI (#101424)

dagitses committed 2 years ago

b5ed606a

Fix typing for `setup_context` in `autograd` (#101464)

lkct committed 2 years ago

e7681b53

Revert "Actually functionalize torch.export (#101433)"

pytorchmergebot committed 2 years ago

eac5f2a8

SymIntify convNd and conv_transposeNd, fix inductor symint handling (#101488)

ezyang committed 2 years ago

b94f143a

[pt2] skip flaky `linalg_householder_product` tests (#101551)

nkaretnikov committed 2 years ago

411ba1c8

[meta] error checking for inplace ops (#101532)

kshitij12345 committed 2 years ago

afea1a9f

Improve rebase message when PR is uptodate (#101504)

malfet committed 2 years ago

54fe828c

BE changes for tryrebase.py (#101503)

malfet committed 2 years ago

20deccf8

Revert "extend serialization for tensor metadata (#99808)"

pytorchmergebot committed 2 years ago

1272cd73

fix a typo in common_device_type.py (#101485)

Bin1024 committed 2 years ago

3f87c04c

Docs: update default device description (#101283)

snorfyang committed 2 years ago

2a3e45a2

[DTensor][2/N] add DTensor constructor function: empty (#101022)

XilunWu committed 2 years ago

010763be

[DTensor][1/N] add DTensor constructor function: ones (#100933)

XilunWu committed 2 years ago

5cc361c7

log inductor compilation time to scuba (#101317)

dshi7 committed 2 years ago

2af7df62

Actually functionalize torch.export (#101433)

tugsbayasgalan committed 2 years ago

eec752ed

Add top level function to check if running with deploy (#101420)

albanD committed 2 years ago

59dff013

Add missing `torch.distributed.ReduceOp.AVG` in type stubs (#101534)

XuehaiPan committed 2 years ago

05f62508

run buildifier on WORKSPACE (#101411)

dagitses committed 2 years ago

47d31364

Type conversion between float/complex dtypes (#97935)

gautierronan committed 2 years ago

ff3f1961

[inductor] erfc: lowering (#101416)

kshitij12345 committed 2 years ago

2b2a717f

Update llama to failing (#101565)

ezyang committed 2 years ago

23d1cc38

[fx] Better replacements finder in subgraph rewriter (#100556)

angelayi committed 2 years ago

9e023e18

[reland][CustomOp] Add Dispatcher error callback (#101452)

zou3519 committed 2 years ago

6bc0f4a4

[reland][custom_op] Change the python type that maps to ListType in schema (#101451)

zou3519 committed 2 years ago

c8be493d

[reland] Cleanup custom op library after each custom_op test (#101450)

zou3519 committed 2 years ago

4f8cbaa1

Older