strisunshinewentingwang/pytorch

Pull Requests Commits

[FSDP] Ignore buffer type casting in ignored modules (#106766)

weifengpy committed 1 year ago

4bc846c1

[quant][pt2e] Add reference representation for quantized conv2d (#105784)

jerryzh168 committed 1 year ago

97ce979e

Attempt to pin opencv-python version (#106900)

ezyang committed 1 year ago

02e44153

Include fused nodes' debug_str in FusedSchedulerNode::debug_str_extra (#106356)

xmfan committed 1 year ago

787d5259

[dynamo] Readability - Rename name to get_frame_name (#106880)

anijain2305 committed 1 year ago

77acb04a

[dynamo] use cache size to detect recompilation (#106878)

anijain2305 committed 1 year ago

8aca7243

Add F8 BLAS data types conversion (#106843)

malfet committed 1 year ago

c2ddb71a

Adding release compatibility matrix for release 2.1 (#106891)

atalman committed 1 year ago

0b880075

[aarch64] Add PT Docker build image for aarch64 (#106881)

Mike Schneider committed 1 year ago

861ae399

Reordering tests experiment (#106347)

clee2000 committed 1 year ago

7dfab082

Make InternalModel and Resnet work with rexportable flow (#106676)

tugsbayasgalan committed 1 year ago

a44c072c

Avoid subtracting by sys.maxsize when something is bounded below by -sys.maxsize - 1 (#106716)

ezyang committed 1 year ago

8ea13a95

Update torchbench.txt (#106761)

msaroufim committed 1 year ago

1b32ac3c

Remove unused _add_runtime_assertions (#106759)

gmagogsfm committed 1 year ago

47014883

[quant][pt2e] Move annotate functions in XNNPACKQuantizer to utils (#106642)

jerryzh168 committed 1 year ago

e1a17806

[pt2] add Python meta for `triangular_solve` (#106682)

nkaretnikov committed 1 year ago

467a2e63

Change dropout of device Privateuse1 to fused kernel (#106774)

lyriexs666 committed 1 year ago

318fcc5e

[FSDP][Easy] `zeros` -> `empty` for immediately freed tensors (#106857)

awgu committed 1 year ago

6f036c96

Add some const to `IndexKernel.cu` (#106809)

r-barnes committed 1 year ago

a0c0666f

Reenable `torch._int_mm` testing on newer CUDAs (#106840)

malfet committed 1 year ago

387e3b04

[BE] Add optional `t` param to `CuBlasLtMatrixLayout` (#106841)

malfet committed 1 year ago

046d6178

Revert "[core][pruning][feature] cuSPARSELt kernels and ops (#102133)"

pytorchmergebot committed 1 year ago

fe594ab3

[inductor] Switch inductor_prims._bucketize over to aten.bucketize (#106658)

davidberard98 committed 1 year ago

387f1ab1

Enable mypy checking in compile_fx.py (#105830)

chenyang78 committed 1 year ago

40a15b50

add xpu support for foreach kernels (#106021)

weishi-deng committed 1 year ago

088e3166

Revert "conv-bn folding in low precision (#106576)"

pytorchmergebot committed 1 year ago

dc7ec4c8

Revert "inductor: remove conv_bn folding from pre_grad pass (#106686)"

pytorchmergebot committed 1 year ago

0ce103a0

[vision hash update] update the pinned vision hash (#106832)

pytorchupdatebot committed 1 year ago

c876afea

export torch/csrc/dynamo/*.h (#106757)

heidongxianhua committed 1 year ago

66914131

Update opset version warning text (#106830)

justinchuby committed 1 year ago

e1e6bbd8

Newer Older