strisunshinewentingwang/pytorch

Pull Requests Commits

[Expanded Weights] add ability to not specify batch size (#80944)

samdow committed 2 years ago

3d74fd48

[Expanded Weights] fix conv3d (#80943)

samdow committed 2 years ago

5973cbe6

[Expanded Weights] fix layer norm (#80895)

samdow committed 2 years ago

2bcbea1f

Revert "[Codemod][Format buck files with arc lint] caffe2/third_party (#81441)"

pytorchmergebot committed 2 years ago

74080044

Revert "Add should_traverse_fn to torch.fx.node.map_aggregate (#81510)"

pytorchmergebot committed 2 years ago

fde1107f

[ao] Updated dict keys of detectors to have consistent naming scheme (#81587)

vspenubarthi committed 2 years ago

e907a8d9

[torchgen] support codegen'd C++ API for a mixture of namespaces (#81581)

larryliu0820 committed 2 years ago

9f873ed7

Remove `eval` from `torch.storage._TypedStorage.__new__` (#81679)

kurtamohler committed 2 years ago

8367fd9d

[JIT] Add may_contains_alias function in SchemaInfo class (#81444)

goldenxuett committed 2 years ago

a6e716cf

[JIT] Fix double wildcard edge case for may_alias in SchemaInfo and improve formatting (#81439)

goldenxuett committed 2 years ago

47cdab66

[functorch hash update] update the pinned functorch hash (#81674)

pytorchmergebot committed 2 years ago

18a0737f

Refactored prim utils into _prims_utils folder (#81088)

Chillee committed 2 years ago

80231d0a

Revert "Adding fsdp fp16 and bf16 hooks (#80557)"

pytorchmergebot committed 2 years ago

a8f4011e

[quant] Implement forward and backward autograd functions for fake quantize (#81438)

asl3 committed 2 years ago

36801853

[LT] Add a new backend interface [DUP of the original] (#81662)

Krovatkin committed 2 years ago

4aac42cc

torch.distribution examples rendering issue (#81611)

mamaj committed 2 years ago

29a99287

[ao] Standardized InputWeightEqualizationDetector output to single level (#81586)

vspenubarthi committed 2 years ago

8a3f88b5

GIL acquire needed in ValueCache::trimPrefixes (#81061)

slgong-fb committed 2 years ago

72de816f

shard `trunk / linux-bionic-cuda10.2-py3.9-gcc7 / test (slow` to 2 (#81569)

clee2000 committed 2 years ago

d80bce7a

[vulkan] Update allocation parameters for uniform buffers (#81636)

SS-JIA committed 2 years ago

af33c244

[PyTorch] Support norm_first in nn.TransformerEncoderLayer fast path (#78269)

swolchok committed 2 years ago

7230b67c

Adding fsdp fp16 and bf16 hooks (#80557)

Olga Andreeva committed 2 years ago

f7d68284

[ao] Standardize PerChannelDetector Output to be single level (#81585)

vspenubarthi committed 2 years ago

2ddb722b

[Expanded Weights] fix loss reduction (#80892)

samdow committed 2 years ago

799bc645

[build] Add `fbsource` keyword to `pt_operator_libarary` (#81298)

larryliu0820 committed 2 years ago

71e16f9e

Use slow but safe formula for prod_backward (#81617)

ezyang committed 2 years ago

84c8a9f8

Revert "[CI] Move CUDA-11.6 to Python-3.10 configuration (#81233)"

pytorchmergebot committed 2 years ago

c9648580

[Codemod][Format buck files with arc lint] caffe2/third_party (#81441)

jdonald committed 2 years ago

1233c3c2

[DataPipe] Fix fork/unzip with a single child (#81502)

ejguan committed 2 years ago

ccbf04dd

make clamp decomps use torch.* calls, move clamp_min/clamp_max to refs (#81619)

ngimel committed 2 years ago

50d205c5

Newer Older