strisunshinewentingwang/pytorch

Pull Requests Commits

also in run_test

clee2000 committed 2 years ago

f6d6013f

remove non existent files

clee2000 committed 2 years ago

02f61d29

linemaps for inductor: python 3.9 and lower doesn't have bisect key argument (#97369)

zdevito committed 2 years ago

116a4f23

[inductor] use real data for cudagraphify (#97363)

zdevito committed 2 years ago

3303f544

[EASY] Do hook sizes check with SymInt (#97362)

ezyang committed 2 years ago

a1edf5f6

Update xla pin merge rule for python3.8 (#97371)

clee2000 committed 2 years ago

5425191f

[ci] Onnx test 3->2 shards (#97383)

clee2000 committed 2 years ago

bc268284

[WIP][Stronghold] Integrate python API BC-linter from test-infra (#96977)

izaitsevfb committed 2 years ago

191a2322

Upload failed and rerun tests (#97304)

ZainRizvi committed 2 years ago

712bd9ae

[aot autograd] refactor to make functionalization self-contained (#96341)

bdhirsh committed 2 years ago

545abc29

Fix missing dynamo cache lookup registration in profiler.profiler (#97305)

wconstab committed 2 years ago

e8a722b9

Add an issue template to disable CI jobs (#97045)

huydhn committed 2 years ago

ec54f186

[10/N] Remove ST init ops (#96985)

fduwjj committed 2 years ago

5cc2e4d7

rename to need_attn_weights to match elsewhere (#97102)

redwrasse committed 2 years ago

11114ab8

Make early stop the default for checkpoint and expose a way to disable (#96866)

soulitzer committed 2 years ago

7a8b6913

[9/N] Remove ST multiple ops (#96989)

fduwjj committed 2 years ago

546835c4

[prims] Fix schema of minimum_value for a primitive operation (#97327)

kiszk committed 2 years ago

5d5f43ab

Add missing __main__ in two unittests (#97302)

ppwwyyxx committed 2 years ago

726fc366

Add `as_strided_` to tensor docs (#97300)

kiersten-stokes committed 2 years ago

28929b12

Revert "DCE inference graphs too (#97275)"

pytorchmergebot committed 2 years ago

a7856e18

Remove stack trace captures from import (#97274)

jeffdunn committed 2 years ago

d779dadd

Dont increment generation if forward of backward exists, and warning on deallocation of live tensors (#97168)

eellison committed 2 years ago

9c144bc4

[inductor] Rewrite convolution triton templates (#95556)

jansel committed 2 years ago

9370f253

[CI] Add a missing dtype flag in nightly perf run (#97357)

desertfire committed 2 years ago

da96ae23

Revert "FIX make sure we import the correct object from multiprocessing (#81862)"

pytorchmergebot committed 2 years ago

73b7702b

move caffe2/proto/ to its own Bazel package (#97324)

mikey dagitses committed 2 years ago

6273c0af

remove dead torch_pb.h library (#97323)

mikey dagitses committed 2 years ago

364d92f9

[BE][docs]Improve and update checkpoint documentation (#96862)

soulitzer committed 2 years ago

89d116d9

Fixed broken link to troubleshooting.html docs page (#97330)

vfdev-5 committed 2 years ago

0f424f7f

[JIT] Partially support ForwardRef type annotations for NamedTuple attributes (#96933)

davidberard98 committed 2 years ago

a133b508

Older