strisunshinewentingwang/pytorch

Pull Requests Commits

[ONNX] Fix `check_training_mode` in symbolic_helper (#78376)

justinchuby committed 2 years ago

299fbbcc

Generate CUDAConfig.h only for CUDA builds (#78218)

malfet committed 2 years ago

dfd78bf4

[static-runtime] out variant for aten::max (#78271)

tenpercent committed 2 years ago

2679755b

[GH1] Switch trymerge concurrency to be PR-based (#78296)

janeyx99 committed 2 years ago

ac031e13

Remove mentions of py3.5-3.6 (#78318)

kit1980 committed 2 years ago

26852d6f

fix missing dim out of range check for logcumsumexp_cuda with empty source tensor

yuguo68 committed 2 years ago

6ee072a3

Chebyshev polynomial of the first kind (#78196)

0x00b1 committed 2 years ago

029bbe49

[primTorch] refs: margin_ranking_loss, hinge_embedding_loss (#78057)

kshitij12345 committed 2 years ago

8bd8f628

[docs][nn] conv: complex support note (#78351)

kshitij12345 committed 2 years ago

4e1f41f6

[primTorch] Elementwise Binary Ops I (#78023)

Aidyn-A committed 2 years ago

31016eb8

Migrate off xenial gcc5.4 (#78137)

kit1980 committed 2 years ago

e840b82b

[GHF] Add spaces between words in merge fail reasons (#78378)

malfet committed 2 years ago

d2d1d78f

[ONNX] Extract export verification as standalone api from unittest

BowenBao committed 2 years ago

483bb4f0

[Docker] Pin protobuf to 3.20.1 (#78369)

malfet committed 2 years ago

6b4ffa14

Conv BN folding data type issue when conv has no bias (#78241)

leslie-fang-intel committed 2 years ago

1a41cd8f

fixing invalid upstream for rebase to viable/strict (#78371)

swang392 committed 2 years ago

dde56ca3

MPS: add ranked tensors for addcmul ops instead of constants and update version_check (#78354)

kulinseth committed 2 years ago

45462baf

[ci] remove ciflow/all (#78317)

suo committed 2 years ago

d98a8148

[primTorch] Rename is_finite->isfinite (#78211)

zasdfgbnm committed 2 years ago

5ecd30e8

add special handling for resize_() in functionalization pass

bdhirsh committed 2 years ago

92229adf

functionalization: remove some unnecessary view_copies in inplace views

bdhirsh committed 2 years ago

e9c54ae1

move Functionalize dispatch key closer to backends

bdhirsh committed 2 years ago

7ff091fc

make block_diag composite compliant

bdhirsh committed 2 years ago

5cc258ec

[LT] Add IR resuing support for manually-implemented ops

desertfire committed 2 years ago

29189d2b

[docs] Move a sentence from `nn.Transformer` to `nn.TransformerEncoder` (#78337)

ayaka14732 committed 2 years ago

91a4fe07

Fix coreml ios workflow (#78356)

malfet committed 2 years ago

a4723d5a

`BCE` loss: forward-over-reverse AD support (#77852)

nikitaved committed 2 years ago

3924d56f

Revert "MPS: add ranked tensors for addcmul ops instead of constants and update version_check (#78312)"

pytorchmergebot committed 2 years ago

5ef9108a

Hackily use addSkip to track flakiness in common_utils.py (#78292)

janeyx99 committed 2 years ago

50930604

Revert "Revert MPS changes (#78335)"

pytorchmergebot committed 2 years ago

fa54eb1f

Newer Older