strisunshinewentingwang/pytorch

Pull Requests Commits

Revert "[dynamo] Activation checkpointing as higher order op (#101028)"

pytorchmergebot committed 1 year ago

d0db7d62

Revert "[c10d] Bridge c10d and gloo stores. (#100384)"

pytorchmergebot committed 1 year ago

13383f45

Revert "[caffe2/tools/autograd] Fix non-determinism in code gen (#101287)"

pytorchmergebot committed 1 year ago

2341bd69

Add pattern to merge/simplify split-cat (#100713)

devashishshankar committed 1 year ago

3e1c8168

[MPS] Add support for Custom Kernels (#100661)

razarmehr committed 1 year ago

721b144f

Update torchbench pin (#101365)

ezyang committed 1 year ago

f48718f7

[Profiler] Fix HTML plot output for profiler export_memory_timeline (#101316)

aaronenyeshi committed 1 year ago

e35323d6

Fixed bug in interpolate when interpolation size is larger than max (#101403)

vfdev-5 committed 1 year ago

a8ea4178

Fix some NVCC warnings (Part 2) (#101383)

cyyever committed 1 year ago

a9413564

ASAN: fix use-after-free (#101400)

AlekseiNikiforovIBM committed 1 year ago

effe1425

Revert "[fx] change from #users to num_users in graph printout (#101140)"

pytorchmergebot committed 1 year ago

66eef314

[BE]: Cleanup deprecated stdlib imports (UP006,UP035) (#101361)

Skylion007 committed 1 year ago

616208b4

put third_party/ittapi/ in .bazelignore (#101364)

dagitses committed 1 year ago

1b7d8750

Revert "implement a function to convert a storage to copy-on-write (#100819)"

pytorchmergebot committed 1 year ago

cca31f17

Re enable AutogradNotImplementedFallback on Windows (#101062)

mantaionut committed 1 year ago

bfb2888b

fix precision error in constraint solver (#101307)

avikchaudhuri committed 1 year ago

9b6ccde0

Revert "[inductor] fix incorrect strides in copy() decomp, fix hf_LongFormer + hf_BigBird errors (#100115)"

pytorchmergebot committed 1 year ago

87f9160b

[pt2] add meta for `linalg_ldl_factor_ex` (#101362)

nkaretnikov committed 1 year ago

7dd8e088

[pt2] add meta and `SymInt` support for `linalg_householder_product` (#101315)

nkaretnikov committed 1 year ago

a8964d63

Inductor cpp wrapper: fix FallbackKernel support (#100788)

chunyuan-w committed 1 year ago

cc54da48

Fix Math Typesetting for torch.linalg.matrix_exp (#101363)

JJGO committed 1 year ago

72908e76

Make missing model import error marginally better (#101221)

ezyang committed 1 year ago

fcf2fb27

Refactor after_dynamo to have a CLI interface too. (#101220)

ezyang committed 1 year ago

96487d0d

Fix torch.utils._traceback on Python 3.11 (#101277)

ezyang committed 1 year ago

9ba64cba

[BE]: Bugfix functorch and some generic typing improvements (#101337)

Skylion007 committed 1 year ago

dfe484a3

[dynamo] Graph break on ops having inplace_view tag (#100787)

blzheng committed 1 year ago

65412f95

[dtensor] Relax condition for _split_tensor() (#101218)

wz337 committed 1 year ago

568db1b4

[vision hash update] update the pinned vision hash (#101347)

pytorchmergebot committed 1 year ago

674e52b0

[transformer benchmark] fix in sdp_bwd for scaled_dot_product_attention return type (#101341)

jspark1105 committed 1 year ago

8876c0b2

Update doc strings to make description of is_causal consistent for nn.Transformer and nn.MHA (#101089)

mikekgfb committed 1 year ago

2361f7f0

Newer Older