strisunshinewentingwang/pytorch

Pull Requests Commits

[Dynamo] modify dynamo ipex backend (#94169)

jiayisunx committed 1 year ago

5fe72b87

[MPS] Fix crashes in several backward ops (#94343)

razarmehr committed 1 year ago

877482eb

[vision hash update] update the pinned vision hash (#94358)

pytorchmergebot committed 1 year ago

61ecaf1d

Cleanup hung Windows processes (#94357)

huydhn committed 1 year ago

5f25c083

Tiny unimplemented improvements (#94150)

voznesenskym committed 1 year ago

68b35017

Remove overly strict assert, add test (#94151)

voznesenskym committed 1 year ago

b191a5f7

Check the semantic of loading the mask value (#91755)

EikanWang committed 1 year ago

88ef4739

add torch.autograd._set_view_replay_enabled, use in aot autograd (#92588)

bdhirsh committed 1 year ago

83275d8c

Add benchmarks.py to run all benchmarks, add new file with all torchbench model names (#94146)

voznesenskym committed 1 year ago

333e7713

Simplify CMake CUDNN code (#91676)

cyyever committed 1 year ago

5fa71207

Simplify cmake code (#91546)

cyyever committed 1 year ago

9291f9b9

[MPS] Add MPSAllocatorInterface to access methods of MPSAllocator (#94327)

razarmehr committed 1 year ago

c981b7e5

[inductor] fix cpu implementation of argmax / argmin (#94165)

Valentine233 committed 1 year ago

51b487bf

change the dynamo benchmark timeout as a parameter (#94284)

chuanqi129 committed 1 year ago

94394e56

Handle sympy in split (#94285)

voznesenskym committed 1 year ago

f48b4d88

Apply some safe comprehension optimizations (#94323)

Skylion007 committed 1 year ago

3ce1ebb6

[NestedTensor] Call contiguous in linear backward (#94317)

drisspg committed 1 year ago

bef2483e

[FSDP][optim_state_dict] Returns the initial states of the empty parameters for KeyedOptimizer/NamedOptimizer (#94130)

fegin committed 1 year ago

ab4fe01e

torch.inference_mode: add type hints (#94223)

adamjstewart committed 1 year ago

ec25db77

Test enabling full testing on 3.11 for linux (#94056)

albanD committed 1 year ago

75e04f6d

Use the right run_test for inductor opinfo tests (#94312)

albanD committed 1 year ago

34bbd7af

Add another missing decomp (#94113)

mlazos committed 1 year ago

d16c2c36

[vulkan] Add core graph components (#94222)

SS-JIA committed 1 year ago

6b8eb0eb

[BE]: pyupgrade Python to 3.8 - imports and object inheritance only (#94308)

Skylion007 committed 1 year ago

8fce9a09

Revert "[inductor] fix crash issue when input is a view tensor (#90150)" (#94329)

seemethere committed 1 year ago

567e6152

Add deprecation warning to reduce flag of scatter for Tensor src and redirect to scatter_reduce (#94282)

mikaylagawarecki committed 1 year ago

7b3217e6

[BE]: Apply pyupgrade yield from and unit test alias upgrades (#94309)

Skylion007 committed 1 year ago

748bac87

[easy] Add NestedTensorMeta to parseDispatchKey (#94279)

mikaylagawarecki committed 1 year ago

895d4781

Factor out SYMPY_INTERP (#94307)

ezyang committed 1 year ago

8c835a9e

Add CSR->BSC and CSC->BSR conversions (#93301)

alexsamardzic committed 1 year ago

e1f17b35

Newer Older