strisunshinewentingwang/pytorch

Pull Requests Commits

[BC-breaking] Use ScatterGatherKernel for scatter_reduce (CPU-only) (#74226)

mikaylagawarecki committed 2 years ago

2bfa0184

[static-runtime] optimize empty if blocks at runtime (#74987)

tenpercent committed 2 years ago

11c412a8

Add forward AD for torch.atan2

IvanYashchuk committed 2 years ago

65ed1e35

Add mapping for unsqueeze_n_times (#75043)

Shirong Wu committed 2 years ago

cf66d420

flatten support - C++ only (#1545)

zasdfgbnm committed 2 years ago

Verified 66a04e90

Revert "masked argmin and argmax"

pytorchmergebot committed 2 years ago

dfce8e25

Add non-eager registration to dispatch autogen (#74557)

wconstab committed 2 years ago

b9e535a6

stft: Implement center padding in ATen

peterbell10 committed 2 years ago

c5023ea1

Implement F.pad in ATen

peterbell10 committed 2 years ago

7f051b4d

[TorchArrow][AIBench] Add AIBench Metrics for TorchArrow Inference Benchmark Test (#75035)

Norman Ponte committed 2 years ago

2e8b9c77

Pin cmake version to workaround pytorch build issues

xuzhao9 committed 2 years ago

62c6801e

masked argmin and argmax

george-qi committed 2 years ago

cb687773

[torch.package] add utility for determining where bad modules may come from (#74998)

bradleyhd committed 2 years ago

4568daf5

Fix ir_metadata Python frames func and remove dead code (#74979)

wconstab committed 2 years ago

14affba7

Disallow functions that are in submodules to also be methods

lezcano committed 2 years ago

8d69254d

Changes to support input sequence ID tracking (#70264)

alexmsettle committed 2 years ago

c0a6add7

[FSDP] Register state_dict hooks for FlatParamsWrapper even if params_list is empty (#74860)

fegin committed 2 years ago

fd4ad5d7

Introducing SymInt to Pytorch (for tracing size arithmetic) (master rebase) (#74861)

Krovatkin committed 2 years ago

5177f95d

Reland: "free up dispatch key space (in C++)" (#74963)

bdhirsh committed 2 years ago

1b7d7d93

Revert "Removed python dispatch keys from dispatch key extraction"

pytorchmergebot committed 2 years ago

3d34afa9

Adding versions to flatbuffer schema (#74989)

pavithranrao committed 2 years ago

6905feea

expanded weights: group norm faster rule

samdow committed 2 years ago

72f7193f

[PyTorch Edge] Add Quantized Softmax Op (Naive Implementation) (#75017)

salilsdesai committed 2 years ago

8d7242a1

expanded weights: layer norm faster rule

samdow committed 2 years ago

8b8f3e83

Fix ID counting with rfactor (#1544)

naoyam committed 2 years ago

Verified 7ed57c9a

[ROCm] unskip FFT tests

micmelesse committed 2 years ago

c4321e13

remove unused nn_path from generate_code (#74563)

mikey dagitses committed 2 years ago

60729d02

Use ONNX Runtime 1.11 in CI

garymm committed 2 years ago

1a1f0d4b

Revert D35000703: [WIP][FSDP] Mixed precision enablement

malfet committed 2 years ago

a98d1a5f

Catch overflows in calculating storage byte size

peterbell10 committed 2 years ago

13a3e5c7

Newer Older