strisunshinewentingwang/pytorch

Pull Requests Commits

[fx2trt] Add torch.nn.function.pad support for fx2trt (#67498)

842974287 committed 2 years ago

4a106e41

[nnc] Fixed handling of 0-sized tensors in cat (#67734)

navahgar committed 2 years ago

383c1f51

Fix adaptive_max_pool2d for channels-last on CUDA (#67697)

xwang233 committed 2 years ago

31cf3d6a

[TensorExpr] Add lowering for aten::max (reduction). (#66519)

Mikhail Zolotukhin committed 2 years ago

ff5c61a7

[TensorExpr] Add lowering for aten::embedding. (#66518)

Mikhail Zolotukhin committed 2 years ago

00afe9ba

[TensorExpr] Add lowering for aten::conv1d. (#66517)

Mikhail Zolotukhin committed 2 years ago

008a58d2

[TensorExpr] Fix lowering for aten::softmax for the case when dtype parameter is None. (#66516)

Mikhail Zolotukhin committed 2 years ago

d58ef2bb

Modify "gemm" code to enable access to "sbgemm_" routine in OpenBLAS (#58831)

austinpagan committed 2 years ago

ea4d9838

Split channels_last test cases for tensor conversion OpInfos (#67368)

zou3519 committed 2 years ago

05d1dcc1

add a quantized hardsigmoid inplace variant (#65740)

vkuzo committed 2 years ago

92a85ecb

ATen | Fix potential crash if `MTLCreateSystemDefaultDevice` return nil (#66859)

mattjgalloway committed 2 years ago

e32d7f75

[PyTorch][Static Runtime] Separate overlap checks for easier debugging (#66637)

swolchok committed 2 years ago

51033649

add jit_trace_module python binding (#67425)

Krovatkin committed 2 years ago

3db536e5

type inputs (#67424)

Krovatkin committed 2 years ago

a8757cdd

add a few convenience helpers to removeAllXXX to Block and Node (#67423)

Krovatkin committed 2 years ago

d3525872

[FSDP] CPU offload resubmit (#67249)

rohan-varma committed 2 years ago

7f3326a6

[NOOP][clangformat][codemod] Enable CLANGFORMAT for caffe2/caffe2/* (#67624)

gandalf2390 committed 2 years ago

06d1be24

[Static Runtime] Add PyTorchPredictor::predict_managed_result to return managed output tensors (#65598)

d1jang committed 2 years ago

e86a5a3a

Raise warning when calling collectives on non-member group objects (#67639)

mrshenli committed 2 years ago

18955d35

[quant][fx] Add support for fused modules in _convert_do_not_use (#67245)

jerryzh168 committed 2 years ago

54241a9c

[BE] [GHA] Use `aws ecr get-login-password` (#67709)

malfet committed 2 years ago

91971dfc

Changed value constraint of exponential dist (#67184)

JAEarly committed 2 years ago

16ee6409

[PG NCCL] Disable NCCL health check (#67668)

rohan-varma committed 2 years ago

885da61d

Remove special FX OpInfo list (#67520)

Chillee committed 2 years ago

0b2f68ea

Remove native_functions.yaml dependency from Sorting.cu (#66621)

peterbell10 committed 2 years ago

96e3d1a7

Remove native_functions.yaml dependency from ScanKernels.cu (#66620)

peterbell10 committed 2 years ago

7deb1726

.github: Migrate iOS workflows to GHA (#67645)

seemethere committed 2 years ago

9e97ccbd

[PyTorch Edge] Use Integer Subtraction (Instead of Float) in Non-FBGEMM Dequantization (#67115)

salilsdesai committed 2 years ago

a8317137

[PyTorch Edge] Parallelize Quantize and Dequantize Tensor (#65845)

salilsdesai committed 2 years ago

23bd3cf5

[PyTorch Edge] Clean up Quantize Tensor code (#66220)

salilsdesai committed 2 years ago

92cfda17

Newer Older