strisunshinewentingwang/pytorch

Pull Requests Commits

disable failing circleci test jobs

clee2000 committed 1 year ago

d0ced64c

[ao] fixing public v private for fx.pattern_utils.py (#86033)

HDCharles committed 1 year ago

d6bfbdf5

[ao] fixing public v private for fx.graph_module.py (#86032)

HDCharles committed 1 year ago

bf0116d1

[ao] fixing public v private for quantization_types (#86031)

HDCharles committed 1 year ago

25476f2e

Use CUTLASS GEMM for NT bmm [OSS-only] (#85894)

cpuhrsch committed 1 year ago

ef58a132

Display unexpected exceptions raised from test_dtypes (#86599)

peterbell10 committed 1 year ago

73c43ce2

Add AutocastHPU support (#84927)

australopitek committed 1 year ago

6be9d9a6

Disable tf32 in functorch transform tests (#86799)

zou3519 committed 1 year ago

553eaaba

[primTorch] Add ref for `triplet_margin_loss`, improve `triplet_margin_with_distance_loss` (#85614)

nkaretnikov committed 1 year ago

d56017a1

Extend torch.cuda.is_available() to attempt an NVML-based CUDA availability assessment when explicitly requested by the user (#85951)

speediedan committed 1 year ago

ce56ee11

Add prims.clone (#86705)

IvanYashchuk committed 1 year ago

cd7c86ea

[BE] Store helper functions C++ for python API parity (#82136)

H-Huang committed 1 year ago

3356d038

[ONNX] Support device().type() string comparison with constant (#86168)

BowenBao committed 1 year ago

cc7ea93c

[ao] fixing public v private for backend_config.native.py (#86030)

HDCharles committed 1 year ago

58542eb2

Added define to fix issue with compatibility with latest Windows SDK (#85408)

vladimir-aubrecht committed 1 year ago

409efeba

Allow PrivateUse1 backends to not have Storage (#86557)

smk2007 committed 1 year ago

f24d174f

use cff standard for citation information (#86200)

pmeier committed 1 year ago

61a58986

[primTorch] decomposition for bucketize (#86366)

Fabio Rocha committed 1 year ago

493ded24

Patching getitem in partitioner (#86713)

jjsjann123 committed 1 year ago

f903f1ab

[primTorch] special: entr, expit (#86592)

khushi-411 committed 1 year ago

23441351

Add type and shape annotation for gm.print_readable() (#86562)

SherlockNoMad committed 1 year ago

a47f93b6

Revert "Backport currently dont work with some models if: (#86510)"

pytorchmergebot committed 1 year ago

e0d6898c

(Re-open) Adds cudaMallocAsync as an alternative backend for the CUDA allocator (#82682)

eqy committed 1 year ago

25725fd6

Add testing on A10G GPU to periodic workflow (#85524)

malfet committed 1 year ago

a216f470

Disable autocast in aot autograd (#86515)

eellison committed 1 year ago

c4f0b93f

Basic SDP benchmark harness (#86729)

cpuhrsch committed 1 year ago

d598290b

Backport currently dont work with some models if: (#86510)

qihqi committed 1 year ago

4bfb7341

Re-enable torchdynamo unit tests (#86658)

desertfire committed 1 year ago

ce48df9e

[MPS] Extend unary ops to int64 (#86615)

malfet committed 1 year ago

692b525b

Revert "Enable max.unary_out (#85926)"

pytorchmergebot committed 1 year ago

f912b585

Older