strisunshinewentingwang/pytorch

Pull Requests Commits

Enable Running Variable Size Row/Col Indices q8gemm Sparse Kernels in QNNPACK (#85244)

salilsdesai committed 1 year ago

6c6a32c2

Update Qnnpack Fully Connected Sparse Op to Store Variable Size Indices (#85243)

salilsdesai committed 1 year ago

4c0e1dc9

Add functorch shard to sm86-periodic workflow (#86820)

malfet committed 1 year ago

1a87c25f

Make `ASGD` & `RProp` differentiable (#86258)

Emilio Castillo committed 1 year ago

cb4867a7

Spread distributed backends among all distributed shards (#86837)

huydhn committed 1 year ago

52249067

Fix typo TORCH_ONLY_METHOD_OPERATORS -> TORCH_ASSERT_ONLY_... (#86661)

peterbell10 committed 1 year ago

48c648d7

[ao] fixing public v private for fx.quantization_types (#86036)

HDCharles committed 1 year ago

67fbd940

[ao] fixing public v private for quantization_patterns.py (#86034)

HDCharles committed 1 year ago

b00cdb5b

[primTorch] special: ndtr, ndtri, log_ndtr, erfcx (#86077)

khushi-411 committed 1 year ago

77d29bce

Fix up cond a bit to make it work w/ fake tensor (#86727)

voznesenskym committed 1 year ago

ea586c05

[easy] Add nested tanh (#86826)

mikaylagawarecki committed 1 year ago

2a751525

Make the data types of output and input consistenst for batchnorm (#84410)

CaoE committed 1 year ago

b79bac0e

[flakybot] add dynamo as platform (#86701)

clee2000 committed 1 year ago

c2f29e75

Allow viable/strict promotion even if periodic or docker-release-builds jobs are failing (#86827)

ZainRizvi committed 1 year ago

94700597

Reland 2 min/max support for SymInt/Floats, finish as_strided/scatter/squeeze() backward symint support (#86797)

albanD committed 1 year ago

66cab524

ci: Just use regular checkout (#86824)

seemethere committed 1 year ago

894c4218

Make `Adadelta`,`Adagrad` & `Adamax` differentiable (#86096)

Emilio Castillo committed 1 year ago

aacb9f3a

[DOC] Use type hints to show annotation in the docs (#79086)

ShawnZhong committed 1 year ago

e552cf10

Improve NestedTensor documentation (#85186)

mikaylagawarecki committed 1 year ago

a77f2a95

Revert distributed test parallelization (#86756)

huydhn committed 1 year ago

be81f3d8

Add hooks for register_buffer/module/parameter (#86148)

antoniojkim committed 1 year ago

09a676f6

Let retried jobs advance viable/strict (#86821)

ZainRizvi committed 1 year ago

c08cbfcc

Update _torch_docs / ldexp (#86721)

vfdev-5 committed 1 year ago

3b266802

[quant][fx] Fix weight_dtype and bias_dtype backend_config checks (#86719)

jerryzh168 committed 1 year ago

363b108e

[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

Newer Older