strisunshinewentingwang/pytorch

Pull Requests Commits

Remove Anjali from CODEOWNERS (#99955)

ezyang committed 1 year ago

650ba572

add Wmissing-prototypes to clang-tidy (#96805)

cyyever committed 1 year ago

dbc7e919

[ROCM] Extend try-catch mechanism for ROCM detection (#99980)

fd0r committed 1 year ago

39ff87c6

[reland][quant][pt2e][refactor] Cleanup the logic for deciding whether to insert observer/fq or not (#99220) (#99767)

jerryzh168 committed 1 year ago

df3455b7

[Testing] flip switch, remove slow path assertions (#99101)

eellison committed 1 year ago

14c3eb7f

[BE] Enable C419 rule for any all shortcircuiting (#99890)

Skylion007 committed 1 year ago

e2a3817d

[inductor] Fix AOTInductor (#99203)

desertfire committed 1 year ago

e43918b9

Get OutOfMemoryError to inherit from RuntimeError (#99786)

gerhean committed 1 year ago

3afa60bf

[inductor] Fix shape padding (#99917)

bertmaher committed 1 year ago

e7157bd0

[pt2] Register meta func to randperm.default (#99593)

xw285cornell committed 1 year ago

cc01568e

[ROCM] Do not send "debug"=True down to triton.compile (#99756)

dllehr-amd committed 1 year ago

8a0badff

Skip some failing dynamic shape models on periodic (#99895)

huydhn committed 1 year ago

9a69634b

inductor: fix issue for conv+binary issue for binary scalar path (#99860)

XiaobingSuper committed 1 year ago

df1ff092

inductor: add fallback test case for hardtanh and leakyrelu fusion pattern (#99859)

XiaobingSuper committed 1 year ago

ed395779

Fix typos and clarify text in tags.yaml (#99954)

ezyang committed 1 year ago

e2cb6bcc

[ONNX] Bump onnx-script version with imported module renaming (#99926)

titaiwangms committed 1 year ago

4cf65462

Return full accuracy status for printing, not abbreviated version (#99894)

ezyang committed 1 year ago

04e8df4d

div16 changes for dyn shapes (#99930)

voznesenskym committed 1 year ago

bfa63bf4

[dynamo] avoid graph break on repeat_interleave.self_int (#99528)

jgong5 committed 1 year ago

e5c9a0fc

Implement the get_device method in the storage base class. (#99818)

fakeYan committed 1 year ago

ecd2c718

[ONNX] Improve diagnostics performance (#99936)

BowenBao committed 1 year ago

e5145329

Add fsspec to PT setup.py (#99768)

wz337 committed 1 year ago

466adab7

Allow persistent reductions in dynamic shapes if last numel is static (#99789)

voznesenskym committed 1 year ago

4be0aa13

yolov3 dynamic training accuracy is fixed (#99896)

ezyang committed 1 year ago

cd617071

inductor: align inductor behavior with eager mode for split_with_sizes (#99702)

XiaobingSuper committed 1 year ago

41069f2f

Use memoized only mode for guard size/stride extraction (#99742)

voznesenskym committed 1 year ago

96ceae3a

Stop marking sequence length as dynamic (#99889)

ezyang committed 1 year ago

0b545bc6

[torchgen] accept scalars for unary `SymInt` arrays (#99921)

nkaretnikov committed 1 year ago

42921fc8

make ATen/native/cuda/NaiveConvolutionTranspose3d.cu data_ptr-correct (#99347)

mikey dagitses committed 1 year ago

1dbecbf9

make ATen/native/cuda/SortStable.cu (#99340)

mikey dagitses committed 1 year ago

4ca44d32

Newer Older