strisunshinewentingwang/pytorch

Pull Requests Commits

Add APoT tensor class and tests

asl3 committed 2 years ago

7390658e

Make lintrunner compatible with M1 (#78628)

malfet committed 2 years ago

d9902779

`index_select` for COO CUDA tensors. (#77551)

nikitaved committed 2 years ago

03cf01bd

Mark more methods of DispatchKeySet as constexpr

zou3519 committed 2 years ago

de5a2320

tools: Add ability to grab release versions

seemethere committed 2 years ago

ffaee661

Use `_all_gather_base` and fuse matmul for sharded linear.

pritamdamania87 committed 2 years ago

44aa4ad8

Fuse row-wise sharded linear matmul to increase perf.

pritamdamania87 committed 2 years ago

effd2709

[coreml] Introducing Quantization (#78108)

mcr229 committed 2 years ago

93d5a722

Revert "Reference implementations for rot90, roll, atleast_1d,2d,3d (#78080)"

pytorchmergebot committed 2 years ago

2d5eac48

Revert "[primTorch] Use the same error message as in ATen for canonicalize_dim (#78541)"

pytorchmergebot committed 2 years ago

6dafefe3

[PyTorch][easy] Fix borrowing from optional in binary_cross_entry_with_logits

swolchok committed 2 years ago

40ada911

[PyTorch] Clean up native transformer implementation

swolchok committed 2 years ago

c944cf87

[DOCS] Add docstring to _get_async_or_non_blocking in _utils.py (#78036)

Justin W. Lin committed 2 years ago

bfc3b955

[primTorch] Use the same error message as in ATen for canonicalize_dim (#78541)

IvanYashchuk committed 2 years ago

c054993b

set data permits requires_grad=True on integer tensor

yuguo68 committed 2 years ago

efdb4192

[Quant][docs] Replace qconfig_dict with QConfigMapping in docs

andrewor14 committed 2 years ago

e41389f8

Support saving Bfloat16 tensors for XLA/HPU (#77534)

sujoysaraswati committed 2 years ago

43c09b5c

Move dynamic RPC tests from common rpc tests to tensorpipe tests

H-Huang committed 2 years ago

e78a7835

Add more impl_nvfuser for prims (#78493)

IvanYashchuk committed 2 years ago

20c32503

MPS: add linespace op (#78570)

kulinseth committed 2 years ago

a3bdafec

Make linalg.vector_norm a proper reduction

lezcano committed 2 years ago

619db911

Add complex_to_float option in ReductionOpInfo

lezcano committed 2 years ago

9b97d5d6

Expose is_out to python

tugsbayasgalan committed 2 years ago

c7e9eea9

Cleanup unused CircleCI binary jobs (#78596)

kit1980 committed 2 years ago

9229749f

Remove Python 2-related code from dataloader (#78594)

kit1980 committed 2 years ago

e8bf3a9c

[S271837] alias torch._utils_internal (#78597)

Eric Huang (AI Platform) committed 2 years ago

0d8c9406

Resolve TODO after Python 2 for custom_fwd (#78592)

kit1980 committed 2 years ago

99244435

Fix perf regression introduced in #70943 (#78588)

zasdfgbnm committed 2 years ago

5bcbad76

[primTorch] Adds broadcast_to, column_stack references (#78416)

jjsjann123 committed 2 years ago

7ea9c6ed

[Vulkan] Add cumsum op (#78554)

yangjijamie committed 2 years ago

05208104

Newer Older