strisunshinewentingwang/pytorch

Pull Requests Commits

split by tag (#108892)

Wenting Wang committed 1 year ago

56fcd16a

[ROCm] Enable Lerp tests for complex32 (#108100)

alugorey committed 1 year ago

ea943448

Forward rank and world size info to Torchbench models when using dynamo runner (#108438)

xmfan committed 1 year ago

54c5f474

replace torch::make_unique with std::make_unique (#108866)

cyyever committed 1 year ago

03e35efb

[HigherOrderOp] Should automatically pop modes (#109157)

yanboliang committed 1 year ago

f03b8abd

[HSDP] Updating HSDP test - test_hsdp_init_with_device_mesh (#109202)

wz337 committed 1 year ago

492a93d1

Refactor `test_foreach.py` (#107869)

crcrpar committed 1 year ago

602413a0

`torch.load`: Replaced multiple one byte read() calls during the `_is_zipfile` check with a single call (#109119)

tringwald committed 1 year ago

f7574ea4

Deprecate torch.cross default behaviour (#108760)

lezcano committed 1 year ago

c382ad47

NT support for gt (#109121)

jbschlosser committed 1 year ago

78cd86c5

[ONNX] Remove deprecated functions (#107208)

wangxiyuan committed 1 year ago

263ca7d6

Add test/onnx_caffe2 to ONNX Exporter merge rule (#109295)

Thiago Crepaldi committed 1 year ago

0edb6167

Add meta and OpInfo for _embedding_bag_dense_backward (#109211)

ezyang committed 1 year ago

fe14e43d

Increase tolerances for baddbmm opinfo test (#109164)

int3 committed 1 year ago

b121e4df

[quant][be] Remove test/quantization/pt2e/test_quantize_pt2e_fx.py (#108925)

jerryzh168 committed 1 year ago

9187559e

Revert "[inductor] Lower masked_scatter on CUDA (#108803)"

pytorchmergebot committed 1 year ago

900288f1

Fix the example in the extending.func.rst (#109279)

FFFrog committed 1 year ago

d4990ad5

[dynamo] implement custom dict variable as a general solution for HF's ModelOutput class (#105044)

weifengpy committed 1 year ago

9021fb8d

[inductor] Lower masked_scatter on CUDA (#108803)

peterbell10 committed 1 year ago

e4036ed7

Revert "[inductor] Add ir.Scan and lower aten.cumsum on CUDA (#106581)"

pytorchmergebot committed 1 year ago

800c6656

Added a flag is_cpu to the AOTInductor runtime (#109300)

bertmaher committed 1 year ago

1b502139

[AOTInductor] Add is_cpu for AOTInductorModelContainer (#109287)

muchulee8 committed 1 year ago

3acccb3a

Revert "add Half support for BatchNorm on CPU (#102070)"

pytorchmergebot committed 1 year ago

b226373d

[dynamo] Add BACKEND_MATCH guard to detect and recompile when backend changes (#107337)

ydwu4 committed 1 year ago

94a54b89

Added test for interpolate nearest exact (#108558)

Anstow committed 1 year ago

9b3f5823

[ROCM] Enable test_fn_fwgrad_..._functional_binary_cross_entropy on ROCM (#109038)

alugorey committed 1 year ago

111b9ef3

Run only one pytest parametrization when generating optest (#108936)

ezyang committed 1 year ago

7f1f5afc

[AOTInductor] Skip pre_grad_passes for exported graph. (#109246)

muchulee8 committed 1 year ago

7f7f6267

add Half support for BatchNorm on CPU (#102070)

CaoE committed 1 year ago

b6a1d3fb

[quant] Remove reference representation rewrite for adaptive_avg_pool2d (#108924)

jerryzh168 committed 1 year ago

41e21898

Older