strisunshinewentingwang/pytorch

Pull Requests Commits

Add device range helper and remove sm86 specific check for memory efficient attention (#102985)

drisspg committed 1 year ago

2800a04a

[export] Remove example custom_object_type to type_reflection_method. (#103015)

zhxchen17 committed 1 year ago

6596cfa4

[ONNX] Add FX exporter MaxPool tests (#102773)

titaiwangms committed 1 year ago

27f4dc6c

Disable fallback for custom kernels (#101131)

zou3519 committed 1 year ago

5b700fc9

[PT2][Quant] Move embedding quantization to osss (#103088)

kimishpatel committed 1 year ago

8e0837cf

[inductor] add a few tests to verify view_to_reshape pass is safe (#103034)

shunting314 committed 1 year ago

bf312f2d

[Dynamo] No graph break for super(MyConv{1/2/3}d, self).forward and super(MyConvTranspose, self).forward (#102509)

weifengpy committed 1 year ago

61736679

torch.compile docs: "Profiling to understand torch.compile performance (#102862)

davidberard98 committed 1 year ago

038955f4

dst_bin_of_end_center is defined twice (#102755)

lonelyeagle committed 1 year ago

62610554

[Profiler] Update Kineto Submodule (#103031)

aaronenyeshi committed 1 year ago

0279d0b6

[FSDP] Fix for optim state dict (#102901)

rohan-varma committed 1 year ago

dfa64fdd

[BE] Use `value_or` (#103065)

malfet committed 1 year ago

2405c59c

Dont run test files that are already run in test_optim (#103017)

clee2000 committed 1 year ago

08c4a442

Fix rocm sharding (#102871)

clee2000 committed 1 year ago

90fd90dd

Add newline before minified repro path (#103083)

ezyang committed 1 year ago

a867e6db

[inductor] fix a numel expr codegen issue (#103005)

desertfire committed 1 year ago

fbbde8df

[inductor] Turn off autotune_cublasLt for cpp_wrapper (#103004)

desertfire committed 1 year ago

49577c7e

[inductor] Support select_algorithm with cpp_wrapper (#103003)

desertfire committed 1 year ago

44fdfd32

[PT2][Quant] Introduce composable quantizer (#102846)

kimishpatel committed 1 year ago

8824101f

Add Wav2Vec2 HuggingFace support (#103009)

ezyang committed 1 year ago

eeb3c621

Add parametrization version of weight_norm (#103001)

ezyang committed 1 year ago

ba962fef

Move CUDA 11.8 CI jobs to CUDA 12.1, CUDA 11.7 jobs to CUDA 11.8 (#102178)

atalman committed 1 year ago

3a38acf1

[pt2] add `SymInt` support for `linalg.tensorsolve` (#102466)

nkaretnikov committed 1 year ago

1fcc67fd

[pt2] add meta for `_linalg_solve_ex` (#102454)

nkaretnikov committed 1 year ago

ec0aa965

[pt2] add meta for `lu_unpack` (#102937)

nkaretnikov committed 1 year ago

4bda4a7e

Add an env PYTORCH_TEST_SKIP_CUDAGRAPH to skip all cuda graph-related unit tests (#103032)

xwang233 committed 1 year ago

39f3514f

Use C++17 [[fallthrough]]; (#102849)

r-barnes committed 1 year ago

b592e675

remove c10::guts::{max,min} (#102952)

cyyever committed 1 year ago

30e27642

[export] Initial serialization v2 (#102707)

angelayi committed 1 year ago

3a385656

Enable uint8/int8 mkldnn/dense tensor conversion (#102965)

leslie-fang-intel committed 1 year ago

d7035ffd

Newer Older