strisunshinewentingwang/pytorch

Pull Requests Commits

Quantization docs: consilidate all API references on a single page (#66198)

vkuzo committed 2 years ago

fe86f0e0

Create a documentation page for `torch.ao.quantization.QConfig` (#66129)

vkuzo committed 2 years ago

7332ed13

Create separate documentation pages for quantization observers and fake_quants (#66125)

vkuzo committed 2 years ago

f0fa3d11

Create a documentation page for FX graph mode quantization APIs (#66122)

vkuzo committed 2 years ago

a89ac313

[AutoAccept][Codemod][FBSourceClangFormatLinter] Daily `arc lint --take CLANGFORMAT`

CodemodService FBSourceClangFormatLinterBot committed 2 years ago

b96c7aea

Remove apparently unnecessary std::remove_cv_t (#66254)

r-barnes committed 2 years ago

109aa135

Disable "-Wignored-qualifiers" for vec256_bfloat16.h (#66279)

r-barnes committed 2 years ago

4cb4d11e

Back out "Revert D30599136: [Pytorch Edge][tracing-based] build tracer in OSS" (#66267)

cccclai committed 2 years ago

3fe5895a

[PyTorch][jit] Fix excess refcounting in TupleType::compare (#66286)

swolchok committed 2 years ago

1763c254

[jit] Don't force refcount bumps from getTypePtr (#66282)

swolchok committed 2 years ago

fb5a80ff

Fix type checking errors in fx/utils.py (#66311)

EshikaShah committed 2 years ago

85b562dd

[hpc infer] fix bench perf number

842974287 committed 2 years ago

e5f6f356

Fix merge conflict in bc tests (#66356)

janeyx99 committed 2 years ago

904fbada

[PyTorch][Static Runtime] Combine ProcessedNode::{native_,}fn_ (#65414)

swolchok committed 2 years ago

5a67ffe0

clean up mypy nit in torch/jit/_recursive.py (#66253)

vkuzo committed 2 years ago

566922bb

Wextra fix for CUDAApplyUtils.cuh (#66323)

r-barnes committed 2 years ago

4a302a30

Revert D31299350: Back out "Revert D31005792: [NCCL] Init dummy NCCL comms in constructor"

janeyx99 committed 2 years ago

0a48f563

Revert D30710710: [Pytorch Edge] Support profiling kineto events from external source

janeyx99 committed 2 years ago

c62ed964

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

Newer Older