strisunshinewentingwang/pytorch

Pull Requests Commits

Replace _baddbmm_mkl_ with cpublas::gemm_batched (#66165)

peterbell10 committed 2 years ago

c957d9fd

Wextra fix 1 for caffe2 (#66272)

r-barnes committed 2 years ago

51835bec

[ao_migration] torch/nn/intrinsic: torch.quantization -> torch.ao.quantization (#65903)

Zafar Takhirov committed 2 years ago

a28b038a

[ao_migration] torch/nn/qat: torch.quantization -> torch.ao.quantization (#65902)

Zafar Takhirov committed 2 years ago

2daae532

[ao_migration] torch/nn/quantizable: torch.quantization -> torch.ao.quantization (#65901)

Zafar Takhirov committed 2 years ago

1a6482ee

[ao_migration] torch/nn/quantized: torch.quantization -> torch.ao.quantization (#65900)

Zafar Takhirov committed 2 years ago

b23709df

Back out "Revert D31005792: [NCCL] Init dummy NCCL comms in constructor" (#65883)

rohan-varma committed 2 years ago

f1f3bd8c

[Pytorch Edge] Support profiling kineto events from external source (#64397)

kimishpatel committed 2 years ago

c1343ff7

Wextra fix for Tensorshape.cpp (#66320)

r-barnes committed 2 years ago

8a02d3e5

Remove cuda/Loops.cuh dependency on native_functions.yaml (#64168)

peterbell10 committed 2 years ago

731cf494

Revert D31445799: [nnc] Use given kernel function name while emitting code

navahgar committed 2 years ago

92ce1885

Revert D31445797: [nnc] Added a cache to use singleton instances of PytorchLLVMJIT for every triple,cpu,attrs combination

navahgar committed 2 years ago

2e6fa026

Revert D31445798: [Static Runtime] Cleanup LLVMCodeGen memory after code gen completes

navahgar committed 2 years ago

097fdcdf

Remove Tensor.h include from TensorIterator.h (#64167)

peterbell10 committed 2 years ago

0be36d79

Migrate THCStorage_resizeBytes to ATen (#65944)

peterbell10 committed 2 years ago

bc1dec9b

Concatting multiple linear layers with same input Tensor (different weight/bias) (#63198)

Gamrix committed 2 years ago

3bad5406

[jit] Implement one-argument AliasDb::mayContainAlias more efficiently (#65177)

swolchok committed 2 years ago

94845fc4

[jit] Add cache for MemoryDAG::collectAllContainedMemoryLocations (#65122)

swolchok committed 2 years ago

c80693f7

[static runtime] Pre-allocate hash tables (#65343)

swolchok committed 2 years ago

3ef69a45

slow_conv3d grad_weight: call gemm directly (#65759)

peterbell10 committed 2 years ago

0020a151

log API usage for fsdp API in PyTorch (#64964)

zhaojuanmao committed 2 years ago

dfb64b32

Revert D31389480: [pytorch][PR] Allow external CUDA streams to be set as current

lw committed 2 years ago

201174cb

[PG Wrapper][BE] Add collective information when monitored barrier error is (#66167)

rohan-varma committed 2 years ago

b72a1782

[PG Wrapper][BE] Make some methods private (#66166)

rohan-varma committed 2 years ago

b5b1d49a

Move intraop_launch_future from Parallel.h (#64166)

peterbell10 committed 2 years ago

0cad2c06

[jit] Reduce refcounting of Types (#65345)

swolchok committed 2 years ago

2d885ab7

[jit] Refcounting spot fixes (#65346)

swolchok committed 2 years ago

1ae468a4

[DataPipe] moving mux IterDataPipe test to the right location (#66277)

NivekT committed 2 years ago

8ebe1a92

[DataPipe] adding test for IterableWrapperIterDataPipe (#66276)

NivekT committed 2 years ago

ed178516

[DataPipe] adding SequenceWrapperMapDataPipe (#66275)

NivekT committed 2 years ago

e808e3d3

Newer Older