strisunshinewentingwang/pytorch

Pull Requests Commits

Minifier fix for non tensor inputs (#83644)

eellison committed 2 years ago

82507ce3

[ROCm] unskip test_jit TestBackendsWithCompiler (#81281)

jeffdaily committed 2 years ago

1f3ef5a2

[MPS] Add native bitwise-not implementation (#83678)

malfet committed 2 years ago

f094113e

CMake: List python source files as codegen dependencies (#83683)

peterbell10 committed 2 years ago

b14df533

[ao] Added Quantization QConfig generation to ModelReport class (#83688)

vspenubarthi committed 2 years ago

5e715be1

Update isDynamic api to align with is_symbolic API (#83415)

miladm committed 2 years ago

72963bba

Check existence of the array ref when tracing resize_ (#81422)

justinchuby committed 2 years ago

04353f78

Coverage for nondeterministic_seeded, respect it in constant prop (#83650)

ezyang committed 2 years ago

91521449

Be more conservative about propagating constants. (#83648)

ezyang committed 2 years ago

24acc315

Address CR comments for "Delete ProxyTensor wrapper subclass" (#83646)

ezyang committed 2 years ago

02581f05

Preserve stack trace for backward nodes over AOTAutograd (#83558)

SherlockNoMad committed 2 years ago

a7baad04

[functorch] add linalg solve batch rule (#82814)

samdow committed 2 years ago

e2e71c1f

[MPS] Fix torch.full for uint8 (#83697)

malfet committed 2 years ago

ff533b1e

Fix and improve the efficiency of the backward of xlog* functions. (#82713)

lezcano committed 2 years ago

88d3acd6

[primTorch] Refs for pdist, triu and related ops (#82819)

Fabio Rocha committed 2 years ago

9e156082

[DataLoader] Move loop content into a function to ensure we don't preserve anything (#83595)

albanD committed 2 years ago

38348362

Add remaining device types in the pybinded DeviceType enum (#83676)

H-Huang committed 2 years ago

23d22724

Revert "Remove conj kernels for real dtypes (#80374)"

pytorchmergebot committed 2 years ago

46ba9f2e

[dist.cp] Introduce LoadPlanner and SavePlanner extensibility API. (#83419)

Rodrigo Kumpera committed 2 years ago

d11d3dd0

[functorch] reclassify svd as an allowed failure; add test (#83612)

zou3519 committed 2 years ago

4a033be4

[functorch] add some vmap+jvp inplace+view tests (#83178)

zou3519 committed 2 years ago

601aca2a

[functorch] relax as_strided batching rule (#83597)

zou3519 committed 2 years ago

d84dc589

[functorch] annotate test_jvpvjp (#83530)

zou3519 committed 2 years ago

69728d7d

[ONNX] Update the default opset to version 14 (#83284)

justinchuby committed 2 years ago

e4f74f08

Added communication hook for sharded cases (#83254)

Olga Andreeva committed 2 years ago

f204afc2

[quant][ao_migration] `torch.nn.quantized.functional` → `torch.ao.nn.quantized.functional` (#78712)

zaf committed 2 years ago

78c8a0d7

[FSDP] Implement sharded_optim_state_dict and flatten_sharded_optim_state_dict. (#77628)

fegin committed 2 years ago

3e1fc85b

Handle python frame is empty in GetPythonFrames (#83643)

JackCaoG committed 2 years ago

cd0ab154

Release the GIL when munmap'ing tensors - fixes #77139 (#83623)

cchan committed 2 years ago

abcf0119

Revert "fixing define_constant pybind signature to match std::complex scalar (#83645)"

pytorchmergebot committed 2 years ago

f84e087d

Newer Older