strisunshinewentingwang/pytorch

Pull Requests Commits

Fix the Requirement of CMake Version (#106254)

FFFrog committed 1 year ago

936333fd

Fix reference cycle in our test suite (#106328)

zdevito committed 1 year ago

8ee0b179

fix torch.norm for custom device (#106198)

heidongxianhua committed 1 year ago

30442c03

Extend Inductor to support the third-party backend (#100706)

EikanWang committed 1 year ago

05bd24bb

correct spelling mistake (#106309)

CCCCarpediem committed 1 year ago

850ad541

Use `isinstance(foreach_arg.type, ListType)` for correctness (#106428)

crcrpar committed 1 year ago

04f20bb2

InductorCpp: Fix "call to constructor is ambiguous" error (#106418)

malfet committed 1 year ago

92cac6bf

Support is_mtia() (#106396)

egienvalue committed 1 year ago

17a31416

[vision hash update] update the pinned vision hash (#106434)

pytorchupdatebot committed 1 year ago

4c3e1371

[MPS] Fix MPS clamp issue with different dtypes between input and min/max tensors (#105747)

DenisVieriu97 committed 1 year ago

d1a2aa19

Remove duplicate ops tests in test_quantized_op.py (#106398)

kit1980 committed 1 year ago

af376082

Fix wrong class name in comments (#106419)

kit1980 committed 1 year ago

dd12c4c2

[inductor][easy] Improved warning message for missing OMP on mac (#106241)

davidberard98 committed 1 year ago

c29f8ccc

[pytorch] Suppress C4624 warnings on Windows (#106348)

jdonald committed 1 year ago

e7115dbe

[quant][pt2e][quantizer] Suppoert set_module_name in XNNPACKQuantizer (#106087)

jerryzh168 committed 1 year ago

92a22a80

Add sequence_nr to aot_autograd to map forward ops to their corresponding backward ops (#103129)

alexmsettle committed 1 year ago

9ba0558d

[DTensor]Minor Docstring Update (#106250)

wz337 committed 1 year ago

0cba33e1

exclude internal folders from lint (#106291)

PaliC committed 1 year ago

5ebb18c5

Refactor stack handling to always use TracingContext to populate real stack on exception (#106277)

ezyang committed 1 year ago

76163a56

aot_inductor: properly split code gen compilation command (#105367)

kflu committed 1 year ago

753991b8

[BE] Input check for torch.nn.MultiheadAttention (#106363)

isdanni committed 1 year ago

5e3aca6c

[Benchmarks] Updated CSVs for improvement to visformer_small (#106414)

drisspg committed 1 year ago

ef0576f2

[ROCm] enabling miopen_batch_norm lowering in inductor (#105740)

jataylo committed 1 year ago

40184b28

Add `_foreach_sign` (#106343)

crcrpar committed 1 year ago

7a3503df

Remove skipIfROCm from test_cuda_repro.py (#106416)

kit1980 committed 1 year ago

b46a89bc

Fix whenRegisteringAutogradKernelWithCatchAllKernel_thenCanCallAutogradKernel (#106023)

FFFrog committed 1 year ago

3ce7abe1

Add cumprod support for device mps (#104688)

Mr4k committed 1 year ago

97e5055a

Expose module method in ExportedProgram (#105575)

tugsbayasgalan committed 1 year ago

fadd0859

[ROCm] enable cudagraph inductor UTs on ROCm (#105662)

jataylo committed 1 year ago

60e65a70

[FSDP][Easy] Move `_FSDPState` attrs to avoid comment confusion (#106392)

awgu committed 1 year ago

506b55fc

Newer Older