strisunshinewentingwang/pytorch

Pull Requests Commits

[optim][radam] fix eps discrepancy for foreach (#92551)

janeyx99 committed 1 year ago

3ba5eae7

Run CI in a new environment (#92378)

malfet committed 1 year ago

97f34e36

[MPS] Fix index_select scalar input with multiple indices (#91064)

qqaatw committed 1 year ago

ccbdf495

Revert "[vmap] unsafe_split : batching rule and OpInfo (#92291)"

pytorchmergebot committed 1 year ago

827e22ec

[primTorch] Remove prims.to_dtype (#92380)

peterbell10 committed 1 year ago

a9f44628

[BE] Get rid of `future` (#92596)

malfet committed 1 year ago

1906eaf2

[reland] Improve hooks ordering behavior (#92559)

soulitzer committed 1 year ago

1bc60c6b

[vmap] unsafe_split : batching rule and OpInfo (#92291)

kshitij12345 committed 1 year ago

0510ae59

Follow up comments of PR #91531 (#92359)

zhaojuanmao committed 1 year ago

0a404fdd

Fix `ShardedTensorMetadata.tensor_properties` for Python 3.11 (#91795)

ConnorBaker committed 1 year ago

20665235

[threaded_pg] fix the comments of MultiThreadTestCase (#92373)

wanchaol committed 1 year ago

06d54b40

[dtensor] delete lagging op db and update op db tests (#92290)

wanchaol committed 1 year ago

997de441

Improve `bsr @ strided` performance in `baddmm` for `bfloat16/half` with Triton kernels. (#88078)

nikitaved committed 1 year ago

8383b5c4

some fixes to get symbolic shapes working through inductor (#92320)

Chillee committed 1 year ago

4f4b62e4

Fix key error formatting and move exc code to exc.py (#92593)

mlazos committed 1 year ago

cac217c8

Make run_dynamic_ci_skips_only.sh more generic (#92581)

ezyang committed 1 year ago

ba682057

[CI] move parallelnative to periodic (experimental) (#92567)

janeyx99 committed 1 year ago

2a7a859d

Remove temporary export skip hack (#92160)

voznesenskym committed 1 year ago

28cb3141

fix promote_constants with ExpandView (#92403)

ngimel committed 1 year ago

ef258642

When nopython=True, Dynamo can't allow graph breaks. (#90970)

wschin committed 1 year ago

bdbd3ed3

Guard on at::Tensor device index (#91779)

voznesenskym committed 1 year ago

eb39d990

[CI] valgrind 3.16.1->3.20.0 (#92552)

malfet committed 1 year ago

388d79cc

Make aot_autograd explicitly error when double backward (#92348)

soulitzer committed 1 year ago

bb779078

[PTD][Oncall] Sync Reorder structure for compatibility with linux-6.0 and gloo submodule for PT (#92568)

fduwjj committed 1 year ago

62eeb7d6

[mergebot] Flatten workflows into jobs, fix bugs (#92097)

clee2000 committed 1 year ago

34353a40

Remove lowering and decompositions of zero_, zero, zeros_like... in favour of their references (#92071)

lezcano committed 1 year ago

8b861544

Fix dynamo.export(aten=True) for condition op (#92361)

SherlockNoMad committed 1 year ago

b5c3b4a3

[optim][asgd] group tensors in foreach to maximize perf (#92364)

janeyx99 committed 1 year ago

c5cb46ec

Fix checking of current mode in PyOperator dispatch (#92357)

angelayi committed 1 year ago

5fdddbbf

Reland 2nd attempt "Add heirachical module names to torchFX graph.node" (#91721)

alexmsettle committed 1 year ago

f8a07ca4

Newer Older