strisunshinewentingwang/pytorch

Pull Requests Commits

make ATen/native/cuda/RreluWithNoise.cu (#99341)

mikey dagitses committed 1 year ago

1b30f588

[pt2] add `SymInt` support for trapezoid ops (#99281)

nkaretnikov committed 1 year ago

fbb0ff10

De-select odd numbered heads from nn.MHA fastpath (#99672)

mikekgfb committed 1 year ago

36e1ae67

[FSDP][Reland] Include duplicate parameters and modules when calling named_parameters and named_modules (#99448)

fegin committed 1 year ago

3de7fd46

Change export tracing_mode default to symbolic (#99877)

ezyang committed 1 year ago

0eb59ad0

Do not assume static by default when exporting (#99554) (#99876)

ezyang committed 1 year ago

73f7459a

[FSDP] Set `NCCL_DESYNC_DEBUG=0` for FSDP unit tests (#99916)

awgu committed 1 year ago

08a8a37f

[spmd] skip gradient copying for fused adam (#99489)

wanchaol committed 1 year ago

855f611b

minor fix to fused adam meta registration (#99436)

wanchaol committed 1 year ago

ca24a962

Improve ProxyTensor tensor_tree list/tuple handling (#99897)

wanchaol committed 1 year ago

ff7d5b62

[fx] Remove replace_literals_with_placeholders (#99728)

angelayi committed 1 year ago

78c2e337

[inductor][non determinism] Disable autotuning when determinisitic mode is ON (#99851)

anijain2305 committed 1 year ago

862d6580

[Lint] Fix wrong docstring for dcp save_state_dict() (#99778)

wz337 committed 1 year ago

7398b565

Fix a minor bug about method generation. (#99704)

fakeYan committed 1 year ago

33fe2dbb

Update pt2-bug-report.yml (#99928)

malfet committed 1 year ago

baf092b8

[c10d] Faster coalescing (#98793)

kwen2501 committed 1 year ago

3a09aa59

[easy] iterate dict with sorted keys for accuracy checking (#99793)

anijain2305 committed 1 year ago

3dcc7b39

[easy] Fix upload test stats after master -> main switch (#99924)

clee2000 committed 1 year ago

2cea2edc

Update MacOS deployment target to 11.0 (#99857)

malfet committed 1 year ago

367d3afd

fix gather issue when index is shape of n by 1 (#99709)

mingfeima committed 1 year ago

4c9d6607

Re-enable dynamic shapes test in dynamo benchmark (#99816)

chuanqi129 committed 1 year ago

e9e5ffe8

Make autocast cache and buffer stealing aware of cudagraph static output tensors (#99368)

eellison committed 1 year ago

d881b297

[CI Testing] Re-enable timm_efficientdet training (#99787)

yanboliang committed 1 year ago

3009c42e

Add support for call_method patterns (#99782)

devashishshankar committed 1 year ago

a1633b17

Don't detach to create parameters in MetaConverter (#99618)

ezyang committed 1 year ago

41280a07

Make new_subgroups avaliable for non-cuda depend backend (#99706)

medivh-xp committed 1 year ago

39590d06

Adding the maintainers approved in 2023Q1 Core Maintainers meeting (#98520)

gottbrath committed 1 year ago

f0e28b1c

Enable ruff in lintrunner (#99785)

justinchuby committed 1 year ago

7d2a18da

[MPS] Add PSO caching for advanced indexing kernels (#99855)

DenisVieriu97 committed 1 year ago

dcd686f4

Improve the precision of abs() and sign() for large values (#99550)

lezcano committed 1 year ago

09b189ed

Newer Older