strisunshinewentingwang/pytorch

Pull Requests Commits

[pytree] Properly register immutable collections (#120036)

khabinov committed 219 days ago

b10ae9e5

Guarantee init cuda before attaching hooks (#120052)

dsjohns2 committed 219 days ago

124c2515

Fix missing right square bracket to match glog format (#119966)

ezyang committed 219 days ago

fbe8e0f9

Add lowering for logcumsumexp (#118753)

amjames committed 219 days ago

9726d7ca

Back out "[pytree] Require serialized_type_name" (#120041)

wilson100hong committed 219 days ago

3f4dd9bf

Add decomp for linalg.cross (#119809)

amjames committed 219 days ago

4625ecb8

Do to convert UnsupportedFakeTensorException into RuntimeError in runNode for proper graph breaking. (#120026)

laithsakka committed 219 days ago

3693d8f4

[DCP][state_dict] Let distributed_state_dict filter out the compiler prefix (#119830)

fegin committed 219 days ago

54025c01

[aot_inductor] move CppWrapperCodeGen into a separate file (#119871)

chenyang78 committed 219 days ago

bc7f3efb

[aot_inductor] move CudaWrapperCodeGen into a separate file (#119870)

chenyang78 committed 219 days ago

78c9b294

Intel GPU Runtime Upstreaming for Device Allocator (#118091)

guangyey committed 219 days ago

8f9f12c0

Add Runtime Constant-Folding function of AOTInductor for AOTInductorModels used internally. (#119823)

muchulee8 committed 219 days ago

b8be8b63

Intel GPU Runtime Upstreaming for Event (#117734)

guangyey committed 219 days ago

4dc75f90

Change native funcol inductor tests to use fake pg (#119104)

yifuwang committed 219 days ago

02fb0435

[Dynamo] Do not create TorchInGraphFunctionVariable for tags (#120005)

oulgen committed 219 days ago

62e5840b

[executorch hash update] update the pinned executorch hash (#119943)

pytorchupdatebot committed 219 days ago

ddde1e4d

Use ARMV8 fconv insns to seepd up scalar fp16<->fp32 (#120012)

malfet committed 219 days ago

4eefe728

Account for inference mode in FakeTensor cache (#119963)

masnesral committed 219 days ago

3e5e8590

fixed flop counter formula for conv transposed backwards pass (#119874)

Chillee committed 219 days ago

8bfc87ce

[FSDP] compile compute and CI with @test_compiled_fsdp (#119933)

weifengpy committed 219 days ago

17c345eb

Setup Nvidia Runtime before Indexer (#119923)

osalpekar committed 219 days ago

c802c501

Add meta registration for _foreach_norm (2nd try) (#119927)

janeyx99 committed 219 days ago

4319735a

[DTensor][Test] Improve math_ops test (#118956)

wz337 committed 219 days ago

707cde9b

[3/N] Replace std::tie with structural binding (#119962)

cyyever committed 219 days ago

94f19fe5

[Dynamo] Support lazy module with namedtuple/dict input (#119972)

yanboliang committed 219 days ago

2a63dd88

Clean up decorators (#119925)

mlazos committed 219 days ago

f9f602fc

Include the scalar tensor auto-transfer in the doc (#119967)

lancerts committed 219 days ago

444c628e

Revert "[export] Change runtime asserts to using assert_scalar (#119608)"

pytorchmergebot committed 219 days ago

47300221

[ROCm] Update triton wheels to ROCm 6.0 (#119765)

jataylo committed 219 days ago

da1df5d7

[inductor][eazy] fix profiler (#119959)

shunting314 committed 219 days ago

3f4f91f2

Newer Older