strisunshinewentingwang/pytorch

Pull Requests Commits

angelayi committed 2 years ago

cea298f2

angelayi committed 2 years ago

dbf18e3b

angelayi committed 2 years ago

6f5686bb

Fix alias issue

angelayi committed 2 years ago

28eba46b

Disable python dispatcher

angelayi committed 2 years ago

9f9116b9

[SPMD] Introduce the cross-iteration graph optimization framework (#94803)

fegin committed 2 years ago

4d9c499d

Remove fake inputs from control flow (#95988)

angelayi committed 2 years ago

5a07c3d3

Add flop formulas for sdpa (#95854)

Chillee committed 2 years ago

9a781ce3

Improve Discoverability of Inductor Optimizations (#95824)

msaroufim committed 2 years ago

7db5f8c7

Fix typo in RELEASE.md and CONTRIBUTING.md (#95965)

kiszk committed 2 years ago

7d02ecfa

Add export docs, improve asserts (#94961)

voznesenskym committed 2 years ago

ac07de4a

Don't use guardless contiguity/stride-like implementations (#95733)

ezyang committed 2 years ago

027ebca4

[EASY] Unindent some blocks (#95967)

ezyang committed 2 years ago

f8b57ba6

Fix: `validate_input_col` for partial functions (#95067)

BeckerFelix committed 2 years ago

e6f3e16d

memory viz: Add colors for categories and a legend (#90587)

zdevito committed 2 years ago

ee438425

[inductor] do not use `ceil` in `arange` ref (#95773)

nkaretnikov committed 2 years ago

c72fbf2e

[inductor] use FP div in CPP expr printer (#95698)

nkaretnikov committed 2 years ago

feffcafe

[DTensor][Shampoo] add _tenso.zero function (#95863)

kesang357 committed 2 years ago

6c061e51

[MTPG] Improve all_reduce and handle bwd thread support (#95524)

Rodrigo Kumpera committed 2 years ago

1d3c394d

[DCP] Add DCP FSDP sharded_state_dict checkpoint example to DCP .rst file (#95517)

wz337 committed 2 years ago

a7698a82

Revert "Don't use guardless contiguity/stride-like implementations (#95733)"

pytorchmergebot committed 2 years ago

4026c621

Run _nvfuser/test_torchscript serially (#95951)

clee2000 committed 2 years ago

7f5f0b36

Revert "[inductor] Add an AOT compilation mode for Inductor CPP backend (#94822)"

pytorchmergebot committed 2 years ago

879400e4

Run more tests through pytest (#95844)

clee2000 committed 2 years ago

d21577f2

Fix formatting (#95906)

Svetlana Karslioglu committed 2 years ago

004bcffc

[Quant][PT2.0] fix issues for rearranging weight observer for decomposed linear (#94296)

Xia-Weiwen committed 2 years ago

f3c25cd3

Triton kernel for bsr @ dense (#94823)

nikitaved committed 2 years ago

d809020f

Move label check failure to mergebot (#94707)

Goldspear committed 2 years ago

88e554b5

[inductor] Add an AOT compilation mode for Inductor CPP backend (#94822)

desertfire committed 2 years ago

73b66098

Inductor: fix crash when indexing an empty tensor by invalid index (#95046)

blzheng committed 2 years ago

3eb8eaa1

Older