strisunshinewentingwang/pytorch

Pull Requests Commits

Update distributed/CONTRIBUTING.md to remove ProcessGroupAgent references and add test instructions

H-Huang committed 2 years ago

24b7142d

[quant][core][better-engineering] Rename files in quantized/cpu directory to conform with non-quantized countertpart filenames

dzdang committed 2 years ago

5874a311

enable with semantics

samdow committed 2 years ago

aa06d052

[PyTorchEdge] Extend Flatbuffer to get mobile_info for NMLML workflows

pavithranrao committed 2 years ago

9b81e817

[complex32, jiterator] cos, sinh, cosh, tanh (#78458)

khushi-411 committed 2 years ago

5fbec86f

Revert "[chalf] where(cpu and cuda), pow(cuda) (#77640)"

pytorchmergebot committed 2 years ago

4bb8db85

Move THPStorage definitions out of `torch/csrc/generic` (#78032)

kurtamohler committed 2 years ago

272193d0

[Profiler] Weaken ordering check during post processing.

pytorchmergebot committed 2 years ago

6a4997e6

Remove call to `.contiguous()` for `local_shard_t`.

pritamdamania87 committed 2 years ago

5aa2ed19

[chalf] warn once on creating a chalf tensor (#78245)

kshitij12345 committed 2 years ago

497ae270

[chalf] where(cpu and cuda), pow(cuda) (#77640)

kshitij12345 committed 2 years ago

3697cf7f

Skip test_fn_gradgrad_linalg_pinv_singular_cuda_complex128

soulitzer committed 2 years ago

cd4ffc86

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

Newer Older