strisunshinewentingwang/pytorch

Pull Requests Commits

[C10d] Add skeleton of LibUV backend. (#105672)

Rodrigo Kumpera committed 1 year ago

2636751f

Add Optimizer state_dict hooks (#105953)

janeyx99 committed 1 year ago

dffa4e14

Add details about ic, broken, flaky, and unstable checks to merge records (#106162)

huydhn committed 1 year ago

4fe407ad

inductor: using dummy input to pack the linear weight for bfloat16 dynamic shape path (#106122)

XiaobingSuper committed 1 year ago

6366ed6e

[inductor] realize boundaries in bucketize() lowering (#106107)

davidberard98 committed 1 year ago

359aa171

No need to wait for pending unstable jobs when merging (#106095)

huydhn committed 1 year ago

32175d79

Support regex flaky rules in trymerge (#106103)

huydhn committed 1 year ago

3b5fb7c0

[ONNX] Support complex in FX exporter (#100554)

titaiwangms committed 1 year ago

eebfb921

[memory snapshot] track context for segments (#106113)

zdevito committed 1 year ago

3e5a52ce

[memory snapshots] removed chained history (#106079)

zdevito committed 1 year ago

45b56476

[vision hash update] update the pinned vision hash (#106182)

pytorchupdatebot committed 1 year ago

73a8544d

[JIT] Fix getting type for subscript assignments. (#106041)

kobrineli committed 1 year ago

32844be3

Update kineto submodule to 465ff4cd7 (#106154)

davidberard98 committed 1 year ago

efeb46e5

sparse.mm.backward: fix for non-contiguous grad values on CPU (#106127)

nikitaved committed 1 year ago

01069ad4

Revert "[quant][pt2e] store scale/zero_point as tensor attributes to support serialization (#105894)"

pytorchmergebot committed 1 year ago

93b2036b

[inductor] Pass to remove pointless clones (#105994)

jansel committed 1 year ago

cb14ff29

[pytorch profiler] fix profiler test for windows (#106156)

briancoutinho committed 1 year ago

e31855d0

Add @penguinwu to distributed codeowners (#105945)

awgu committed 1 year ago

88c400e0

Allow graph breaks in inductor opinfo tests (#105480)

eellison committed 1 year ago

ce633892

[BE] Document optimizer state_dict better, use example (#105958)

janeyx99 committed 1 year ago

ec0ffac3

Add context for warning about batch_first (#106139)

mikekgfb committed 1 year ago

723bc136

Change _dynamo.export to be export(f)(*args, **kwargs) (#106109)

ezyang committed 1 year ago

7b9d250f

[Inductor] Fuse non-foreach ops with foreach ops without iterating over all subnodes (#106008)

mlazos committed 1 year ago

5cbd3fc4

Add pull request target to bc lint (#106065)

clee2000 committed 1 year ago

d4136c90

[easy] improve hint on error message in nn.Module.load_state_dict (#106042)

mikaylagawarecki committed 1 year ago

ca7ece9b

Tag functions to core IR in native_functions.yaml (#105849)

SS-JIA committed 1 year ago

70bc1b0f

Lower batch on cait_m36_384 (#106091)

eellison committed 1 year ago

d9606648

[Easy] remove unneeded sort (#106090)

eellison committed 1 year ago

27ece5fa

[export] Handle the case for no placeholders during in runtime assertion pass. (#106134)

zhxchen17 committed 1 year ago

10f55a2a

[Pytorch][Vulkan] sum.dim_IntList (#105612)

lucylq committed 1 year ago

9ff36c2b

Newer Older