strisunshinewentingwang/pytorch

Pull Requests Commits

[quant][core] Refactor qat conv implementation to use the same _ConvNd as base class (#74505)

jerryzh168 committed 2 years ago

98207aab

guoyejun committed 2 years ago

78cb1abc

Back out "[PyTorch Distributed] Consolidate NCCL_DESYNC_DEBUG and TORCH_DISTRIBUTED_DEBUG=INFO" (#74586)

kwen2501 committed 2 years ago

a15f7834

[mobile] add test model generation script and ios tests

linbinyu committed 2 years ago

c24bc056

Automated submodule update: FBGEMM (#74447)

facebook-github-bot committed 2 years ago

f167a3f9

[pytorch profiler] enable iteration tracking for kineto (#72292)

Brian Coutinho committed 2 years ago

02ba0fa8

skip HPU tensor in TensorIterator (#73343)

gune42 committed 2 years ago

c28df693

Print reason for test skipped in CI (#74451)

tenpercent committed 2 years ago

6a664481

Virtualize `<type>Storage` classes (#66970)

kurtamohler committed 2 years ago

79ddc72b

Make bitwise_left_shift/right_shift scalar variants composite

ezyang committed 2 years ago

2e2cfa2e

Land remaining parts of Torchscript Lazy Tensor backend (#74111)

wconstab committed 2 years ago

3547f208

[SR] Avoid boxing inputs in DictConstruct/ListUnpack (#74250)

Mike Iovine committed 2 years ago

93be0e20

Revert D34805092: Extend _save_for_mobile and _load_for_mobile to support flatbuffer format; Default format is pickle + Change buck targets to support `only pickle` and `pickle + flatbuffer` for migration

malfet committed 2 years ago

c53b3ed2

[ONNX] Adjust `is_train` flag for onnx pass deduplicate initializers

BowenBao committed 2 years ago

144b7de9

[PyTorch] Fix quantized linear_unpack ops not being registered issue (#74526)

larryliu0820 committed 2 years ago

e13229e6

[PyTorch] Fix flatbuffer build error on Android (#74518)

larryliu0820 committed 2 years ago

273e22b6

Prelu OpInfo: change to mostly use positional arg

samdow committed 2 years ago

8ccc484a

Revert "Create jiterator cache dirs recursively"

pytorchmergebot committed 2 years ago

879be6ab

[NNC] call super().setUp() & tearDown() in test_tensorexpr.py (#74504)

davidberard98 committed 2 years ago

3941b1ab

Extend _save_for_mobile and _load_for_mobile to support flatbuffer format; Default format is pickle + Change buck targets to support `only pickle` and `pickle + flatbuffer` for migration (#74209)

pavithranrao committed 2 years ago

284b2b71

docs: expose at::native::unfold (#74224)

polvalente committed 2 years ago

8f55b1d8

Revert D34898108: Process inputs and outputs in fx interpreter

suo committed 2 years ago

bf5e25f3

[GHF] Add test for `GitHubPR.get_last_comment`

malfet committed 2 years ago

ca500f32

Fixes build of CUDAHooks.cpp

zasdfgbnm committed 2 years ago

9a6630e2

Passing explicit pretrained_backbone (#74372)

datumbox committed 2 years ago

ea56b9d9

[GHF] Force dry-run to be passed as named argument

malfet committed 2 years ago

1f12f26f

[GHF] Refactor comment fetching logic

malfet committed 2 years ago

e4beaadc

Process inputs and outputs in fx interpreter (#74242)

842974287 committed 2 years ago

f65594fc

Disable meta device tests.

ezyang committed 2 years ago

65329f4f

Make lazy codegen honor per-operator-headers flag (#74450)

wconstab committed 2 years ago

93f7f588

Newer Older