strisunshinewentingwang/pytorch

Pull Requests Commits

optimize dim reduce performance on norm, argmax and argmin (#72083)

mingfeima committed 2 years ago

75e76944

Remove fx2trt from oss CI (#72595)

wushirong committed 2 years ago

4d01789f

[SR][easy] CPU fuser uses native control flow (#72544)

Mike Iovine committed 2 years ago

c975b928

[tensorexpr] support for fusing autocasting ops (#72478)

davidberard98 committed 2 years ago

2e042957

Automated submodule update: FBGEMM (#72658)

facebook-github-bot committed 2 years ago

dbbdabc3

[DataPipe] Add docstrings for IterDataPipe and MapDataPipe, along with small doc changes for consistency (#72618)

NivekT committed 2 years ago

3e1eff9a

Make test_multiprocessing_spawn.py compatible with pytest (#50408)

zasdfgbnm committed 2 years ago

634427d6

add native quantization support for pixel_shuffle (#68328)

mingfeima committed 2 years ago

0f8a1fc9

[DataPipe] Adding a note about FileLister behavior (#72619)

NivekT committed 2 years ago

b0dd2c2e

Automated submodule update: FBGEMM (#72647)

facebook-github-bot committed 2 years ago

364055b2

small optimizations of a view function (#72626)

Natalia Gimelshein committed 2 years ago

d891b730

Fix tagged build detection for binary builds (#72628)

malfet committed 2 years ago

b6df02bb

Fix refcounting in access of saved for forward attribute (#72627)

albanD committed 2 years ago

ccfafb6e

Gradcheck forward AD respects requires grad but run with requires_grad=False (#72309)

soulitzer committed 2 years ago

91e4f778

Allow caffe2-specific graph transformations for OperatorExportTypes.ONNX_ATEN_FALLBACK when BUILD_CAFFE2 is ON (#67460) (#68490)

BowenBao committed 2 years ago

eb4238fc

[JIT] Cat shape analysis fix for -1 dim (#72616)

Elias Ellison committed 2 years ago

828e36c7

Use opmath_t in cusparseSpMM (#72559)

xwang233 committed 2 years ago

0672cf9e

Use cub 1.15's latest scan-by-key algorithm to replace thrust for Embedding.cu and EmbeddingBag.cu (#66580)

zasdfgbnm committed 2 years ago

9a3b86f6

Ensure all modules in the dict are instance instead of class (#72615)

cccclai committed 2 years ago

239531fb

Fix doc build for release branches (#72567)

malfet committed 2 years ago

c1a4714b

[SR] Make fused_sigrid_transforms work on graph outputs (#71507)

Mike Iovine committed 2 years ago

c76c4912

[DistributedInference] Relax the assertion for uniqueness of blob name across external inputs and outputs (#72492)

singlaiiit committed 2 years ago

2539b6a9

[CircleCI] Deprecate `gpu.medium` class (#72613)

malfet committed 2 years ago

Verified 3b1ef1fd

[jiterator] Move jitted_gpu_kernel into into its own header (#71960)

peterbell10 committed 2 years ago

e235e437

fix minor issues for ATen/ROCm (#71925)

shintaro-iwasaki committed 2 years ago

07824730

Opinfo test for mvlgamma: add epsilon (#72491)

davidberard98 committed 2 years ago

933e0e89

[JIT] Opinfo tests for nnc fusion - retry (#72486)

davidberard98 committed 2 years ago

bbd42c60

Change ParameterList and ParameterDict to be able to contain any kind of objects (#70499)

albanD committed 2 years ago

7035738b

fx quant: add workflow support for torch.matmul quantization (#72444)

vkuzo committed 2 years ago

decc79e5

Automated submodule update: FBGEMM (#72583)

facebook-github-bot committed 2 years ago

979744b6

Newer Older