strisunshinewentingwang/pytorch

Pull Requests Commits

add Quantizer support to IValue (#42438)

Basil Hosmer committed 4 years ago

feeb515a

Revert D22780307: Fix illegal memory acess issue for CUDA versionn of SplitByLengths operator.

Mike Ruberry committed 4 years ago

24e2a8a1

Throw error if `torch.set_deterministic(True)` is called with nondeterministic CuBLAS config (#41377)

kurtamohler committed 4 years ago

df7c0594

enable torch.optim.swa_utils.SWALR (#42574)

vincentqb committed 4 years ago

7221a3d1

Add API to collect output_col_minmax_histogram

Yongbin Gu committed 4 years ago

18a32b80

Add strict mypy type checking and update code_template.py (#42322)

ezyang committed 4 years ago

7c33225c

Freeze dynamic (re)quantizaiton ops into standard ones (#42591)

Yinghai Lu committed 4 years ago

5c5d7a9d

Release the GIL before invokeOperator (#42341)

wconstab committed 4 years ago

6d1e43c5

Fix illegal memory acess issue for CUDA versionn of SplitByLengths operator.

Ren Chen committed 4 years ago

76905527

[c10/cuda] Reorganize device_count() and robustly surface ASAN warnings (#42249)

Dmytro Dzhulgakov committed 4 years ago

06d978a9

[vulkan] VulkanTensor lazy buffer allocation (#42569)

IvanKobzarev committed 4 years ago

27e8dc78

Keep manual_kernel_registration only effective in aten codegen. (#42386)

ailzhang committed 4 years ago

dae94ed0

Add CUDA 11 builds for Windows CI (#42420)

peterjc123 committed 4 years ago

b08347fd

.circleci: Hardcode rocm image to previous tag (#42603)

seemethere committed 4 years ago

db52cd73

Automated submodule update: FBGEMM (#42584)

facebook-github-bot committed 4 years ago

eb8a5fed

Revert D22939119: [TensorExpr] Fix a way we were createing np arrays in tests.

kimishpatel committed 4 years ago

924a1dbe

Unconditinally use typing extensions in jit_internal (#42538)

malfet committed 4 years ago

0cf71eb5

[vulkan] max_pool2d (#41379)

IvanKobzarev committed 4 years ago

b8521688

Hide cudnn symbols in libtorch_cuda.so when statically linking cudnn (#41986)

mcarilli committed 4 years ago

0f358fab

[TensorExpr] Fix a way we were createing np arrays in tests. (#42575)

Mikhail Zolotukhin committed 4 years ago

882ad117

Reenable cusparse SpMM on cuda 10.2 (#42556)

malfet committed 4 years ago

3c7fccc1

handle multiple returns properly in boxing wrappers (#42437)

Basil Hosmer committed 4 years ago

78f4cff8

Reduce the output overhead of OutputColumnMaxHistogramObserver by enabling changing bin_nums, Update the observer_test.py

Yongbin Gu committed 4 years ago

d45e2d3e

Install typing_extensions in PyTorch CI (#42551)

malfet committed 4 years ago

61027a1a

[JIT] Fix torch.jit.is_tracing() (#42486)

Meghan Lele committed 4 years ago

29700c00

[ONNX] Enable lower_tuple pass for custom layer (#41548)

BowenBao committed 4 years ago

afa489de

test: Disable test_strided_grad_layout on ROCM (#42561)

seemethere committed 4 years ago

ccc831ae

Automated submodule update: FBGEMM (#42496)

facebook-github-bot committed 4 years ago

c3e2ee72

Fix the bug in THCTensor_(baddbmm) and ATen's addmm_cuda for strided views input (#42425)

IvanYashchuk committed 4 years ago

b9e68e03

Implement sort for string in aten (#42398)

gmagogsfm committed 4 years ago

317b9d3b

Newer Older