strisunshinewentingwang/pytorch

Pull Requests Commits

[te] Add gflag for fast intrinsic expansion (#49060)

bertmaher committed 3 years ago

b98e62f8

[pe] Add gflags for num_profiled_runs and bailout_depth, laint (#49059)

bertmaher committed 3 years ago

44f33596

[ONNX] Remove usage of isCompleteTensor() in symbolic functions (#48162)

BowenBao committed 3 years ago

e5a98c5a

[PyTorch] Reference to c10::GetCPUAllocator() directly (#49068)

iseeyuan committed 3 years ago

41fd51d7

[numpy] `torch.cosh`: promote integer inputs to float (#48923)

kshitij12345 committed 3 years ago

b3ab25ae

[te] Remove vestigial __init__.py from test/cpp/tensorexpr (#49061)

bertmaher committed 3 years ago

492580b8

Revert D25111515: Extra sampling of record function events

Mike Ruberry committed 3 years ago

9f7fb546

remove redundant sccache wrappers from build.sh scripts (#47944)

jeffdaily committed 3 years ago

73f71784

make validate debug-only in Device copy ctr (#47854)

bdhirsh committed 3 years ago

4b26cafb

Add complex support to broadcast_coalesced (#48686)

Sidney Fletcher committed 3 years ago

71cfb737

Extra sampling of record function events (#48289)

Ilia Cherniavskii committed 3 years ago

09b974c2

[PyTorch] TensorImpl::is_non_overlapping_and_dense_ should default to true (#48625)

swolchok committed 3 years ago

a20d4511

skip cuda test_cholesky_solve_batched_many_batches due to illegal memory access (#48999)

xwang233 committed 3 years ago

a849f382

[ROCm] restore autograd tests (#48431)

jeffdaily committed 3 years ago

e8b00023

Add high level profiling trace for dataloading and optimizer (#47655)

gaoteng-git committed 3 years ago

1c31f762

[quant][graphmode][fx] Add test for ResnetBase (#48939)

jerryzh168 committed 3 years ago

2d9585a6

[pt][quant] Remove contiguous calls in qembeddingbag (#48993)

Hao Lu committed 3 years ago

59a3e766

Annotate torch._tensor_str (#48584)

guilhermeleobas committed 3 years ago

7c0a3e3a

Torch onnx (#48980)

guilhermeleobas committed 3 years ago

34cc77a8

Correctly apply WIN32_LEAN_AND_MEAN to the whole repo (#49025)

peterjc123 committed 3 years ago

5450614c

[quant][fix] Support quantization of ops where input is quantizable (#49027)

supriyar committed 3 years ago

4434c07a

[quant][graphmode][fx] Add MatchAllNode in pattern matching (#48979)

jerryzh168 committed 3 years ago

993ce4b2

Add a pass to fetch attributes of nn.Module to fx.node (#47935)

842974287 committed 3 years ago

107c31f2

[JIT] Allow del statements with multiple targets (#48876)

Meghan Lele committed 3 years ago

3f9ff48e

fx quant: move more functions to utils (#48908)

vkuzo committed 3 years ago

d033e185

fx quant: move qconfig utils to utils file (#48907)

vkuzo committed 3 years ago

2668ea80

fx quant: quick cleanup for model_device (#48906)

vkuzo committed 3 years ago

17e71509

[collect_env] Add candidate paths for nvidia-smi on Windows (#49021)

peterjc123 committed 3 years ago

e538bd66

[CUDAExtension] support all visible cards when building a cudaextension (#48891)

stas00 committed 3 years ago

02b63858

add a unit test for large node error (#48938)

scottxu0730 committed 3 years ago

60004814

Newer Older