strisunshinewentingwang/pytorch

Pull Requests Commits

error on bad input to equality constraint (#107311)

avikchaudhuri committed 1 year ago

95f1591a

Turn on typechecking for _inductor/kernel/conv.py (#106258)

int3 committed 1 year ago

9c9982a0

[inductor] Add ir.WelfordReduction with multiple outputs (#104725)

peterbell10 committed 1 year ago

18b1c290

[DTensor][random] add DTensor constructor: rand (#106535)

XilunWu committed 1 year ago

3699c6ad

[inductor] scatter_reduce - skip .item() in backward if GradMode is not enabled (#107353)

davidberard98 committed 1 year ago

d465d6a8

Turn on typechecking for _inductor/utils.py (#106252)

int3 committed 1 year ago

a815e719

Add scalar conversion using avx instructions for half (#102140)

CaoE committed 1 year ago

1d6a4465

[dynamo][eval_frame] Set destroy_extra_state deleter as part of co_extra (#107117)

anijain2305 committed 1 year ago

b31a357e

[dynamo][eval_frame] Unify cache entry and frame_state on the same co_extra index (#106917)

anijain2305 committed 1 year ago

4608b942

[inductor] Use divisor_override param in in aten.divisor_override lowering (#107401)

masnesral committed 1 year ago

fcd1a0e9

Fix vscode test discovery (#107404)

mlazos committed 1 year ago

2bb59a9a

Fakify leaf of FunctionalTensor (#107062)

ydwu4 committed 1 year ago

6cb0128c

Throw error if `stateless.functional_call` called with `nn.DataParallel` (#107403)

kurtamohler committed 1 year ago

36141de4

[nullability] Suppress -Wnullable-to-nonnull-conversion errors in caffe2 (#107418)

caodoan committed 1 year ago

600f9ef2

Use CUDA 12.1.1 patch version in CI (#107295)

atalman committed 1 year ago

9ca2106e

Fix rst formatting in torch.compiler_troubleshooting.rst (#107360)

apivovarov committed 1 year ago

35b2b3ee

Added xla friendly codepath to single_tensor_adamw (#102858)

aws-murandoo committed 1 year ago

608afe80

[BE] Use allocator to allocate workspace (#107178)

malfet committed 1 year ago

89de0485

Align formula in Impl::General mode with Impl::Contiguous mode in batch_norm_elementwise_cuda op. (#106943)

Shan19900305 committed 1 year ago

3c3874d6

Revert "Add backward check for test_memory_format (#106104)"

pytorchmergebot committed 1 year ago

02bcaf45

Revert "[C10D] Implement new libuv backend for TCPStore. (#105870)"

pytorchmergebot committed 1 year ago

d3f92ca9

Describe the 'float32_matmul_precision' settings in more detail (#107169)

kalekundert committed 1 year ago

26677247

[ONNX] Fix memory leak when exporting models (#107244)

ilyasher committed 1 year ago

2b3917dc

aot_inductor: fix compile returning None if cache hits (#107020)

kflu committed 1 year ago

d8dadb0f

Update test name in multiGPU test (#107397)

fduwjj committed 1 year ago

37eb9699

Make mergebot work with review comments (#107390)

ZainRizvi committed 1 year ago

b9c86c52

[BE] Remove deprecated github gql param and disable inconsistent test (#107385)

ZainRizvi committed 1 year ago

4874b023

Introduce CUDA-only `_scaled_mm` op (#107341)

drisspg committed 1 year ago

8ccfd801

Add backward check for test_memory_format (#106104)

CaoE committed 1 year ago

2e44adb0

Update `generate_opcheck_tests`, also use it to test some internal tests (#107328)

zou3519 committed 1 year ago

c69514cc

Newer Older