strisunshinewentingwang/pytorch

Pull Requests Commits

Fuse nodes with sizes (s0*s1*...,) and (s0, s1, s2, ...) (#120077)

vfdev-5 committed 198 days ago

49d1fd31

[dynamo] Re-dispatch `torch.Tensor.new` into `torch.Tensor.new_empty` method. (#121075)

ysiraichi committed 198 days ago

aa0b0944

[dynamo][guards-cpp-refactor] Prevent duplication of leaf guards (#121164)

anijain2305 committed 198 days ago

e3bd6efe

[dynamo][guards-cpp-refactor] Pass source name for debug ease (#121154)

anijain2305 committed 198 days ago

b6b2d5b0

[dynamo][guards-cpp-refactor] Simplify DictGuardManager by removing KeyValueDictGuardManager (#121147)

anijain2305 committed 198 days ago

52d89d84

[dynamo][guards-cpp-refactor] Add argnames in pybind'ings (#121121)

anijain2305 committed 198 days ago

af7f55ff

[non-strict export] support tensor attribute without other args (#121176)

avikchaudhuri committed 198 days ago

0b9bfcf9

Revert "[XPU][Profiler] Add Logic To The Profiler For Processing XPU-backend Data (#120185)"

pytorchmergebot committed 198 days ago

80879126

torch check the division by zero in batch_norm_update_stats (#120882)

lancerts committed 198 days ago

099ff51d

[BE] Remove `__iniline__` from `__global__` (#121246)

crcrpar committed 198 days ago

2eec0e7c

Capture primitive data type arguments for profiling python_function (#120949)

shengfukevin committed 198 days ago

31bfa599

Batch Norm Consolidation (#116092)

tugsbayasgalan committed 198 days ago

5680f565

__grid__constant is only suported on cuda version >= 11.8 (#121275)

drisspg committed 198 days ago

f72eb5ae

Introduce EphemeralSource for symbols that should be simplified out (#120948)

jbschlosser committed 198 days ago

dad1b765

[FSDP] restore fully_shard after exit from mock.patch (#121058)

weifengpy committed 198 days ago

d968fc44

[cuBLAS][cuBLASLt] Fix expected failures for `int_mm` on `sm75` (turing) (#121277)

eqy committed 198 days ago

8dafc81b

Don't merge qnnpack (#120676)

rodolfozxx committed 198 days ago

ce6a7d56

Add assign argument to torch.Tensor.module_load (#121158)

mikaylagawarecki committed 198 days ago

4b390337

[easy] Fixed requires_grad preservation for nn.Module.load_state_dict(assign=True) (#121157)

mikaylagawarecki committed 198 days ago

27389e03

c10:intrusive_ptr, self assignment (#119275)

DenisYaroshevskiy committed 198 days ago

87a533ed

Fix permuted sum precision issue for lower precision on CPU (#108559)

CaoE committed 198 days ago

412c687e

fix segfault in torch.native_channel_shuffle when input is empty (#121199)

mingfeima committed 198 days ago

34e3f6f3

remove compute capability 3.5 for CUDA 12 (#114930)

njzjz committed 198 days ago

8473cd92

CI: Add aarch64 docker build and ciflow tags (#120931)

snadampal committed 198 days ago

d13ed850

[PyTorch] Split StaticModule out of test_static_runtime (#121028)

swolchok committed 198 days ago

cac36e23

Update docs to point to new sdpa_kernel context manager (#121180)

drisspg committed 198 days ago

f5391dad

[pytorch] Name the main and autograd threads for better debugging (#121170)

valentinandrei committed 198 days ago

8bb3e0b6

[doc] Fix math display in ChannelShuffle doc (#121247)

ssnl committed 199 days ago

24944f67

[ez] Add super() calls in test_custom_ops (#121239)

clee2000 committed 199 days ago

b3a9d677

[Autograd] Improve error for leaf tensors as out argument to fallback (#121089)

peterbell10 committed 199 days ago

34a28f01

Newer Older