microsoft/onnxruntime

Pull Requests Commits

fix BUILD.md typo (#2375)

nihui committed 6 years ago

dde410e0

Another try to stabilize CUDA CI (#2383)

KeDengMS committed 6 years ago

Verified 51571030

Skip layer norm transform (#2350)

liuziyue committed 6 years ago

Verified ffa28125

Fuse Add + Gelu (#2360)

yufenglee committed 6 years ago

Verified 8ed2928d

Layer Norm Fusion Fix (#2379)

liuziyue committed 6 years ago

Verified 4b72fedb

Add opset 11 version of Split to CUDA ops (#2376)

skottmckay committed 6 years ago

Verified 8c733c8d

Fix bug with Slice. Need to pass in flattened input dimensions so the initial offset into the input is calculated correctly. (#2372)

skottmckay committed 6 years ago

Verified c0d23d5f

Extend OneHot CPU kernel to support more types (#2311)

KeDengMS committed 6 years ago

Verified 9e26f4de

update output size calculation for resize (#2366)

askhade committed 6 years ago

Verified 437772d5

Fix a bug in TLS refcount that may destabilized CUDA CI (#2374)

KeDengMS committed 6 years ago

Verified 192dcfaa

test bidaf with nuphar for avx target (#2370)

yangchen-MS committed 6 years ago

Verified 41b9f01e

Add Tracelogging for profiling (#1639)

Changming Sun committed 6 years ago

Verified fc6773a6

fix builds enabling onnxruntime_DEBUG_NODE_INPUTS_OUTPUTS (#2369)

jywu-msft committed 6 years ago

Verified 0c6e9f94

Add helper to create output to minimize binary size. (#2365)

skottmckay committed 6 years ago

Verified 53ed36a3

Direct use python numpy array's memory if already contiguous. (#2355)

zhanghuanrong committed 6 years ago

Verified aa37e2de

Implement cuda nonzero op. (#2056)

zhanghuanrong committed 6 years ago

Verified ed6da0d1

Openvino EP R3.1 onnxrt server (#2357)

avidiyal committed 6 years ago

3d3cf0e1

Fix/test dim value of 0 handling in a couple of places (#2337)

skottmckay committed 6 years ago

Verified 599d72a9

Introduce PrimitiveType into a Type System along with an integer constant (#2307)

yuslepukhin committed 6 years ago

Verified 25b3c516

Set ElementType to String type of node metadata, instead of byte[] (#2348)

jignparm committed 6 years ago

Verified fa30b1e7

Cuda Reverse Sequence Op, maping types of same size using same template function. (#2281)

zhanghuanrong committed 6 years ago

Verified 7fcd7523

Nuget pipeline changes (#2305)

Changming Sun committed 6 years ago

Verified 080a0a31

Remove unused initializer from GraphProto as well as name_to_initial_tensor_ in CleanUnusedInitializers. (#2320)

skottmckay committed 6 years ago

Verified 5a3ea746

implement CPU contrib OP Attention (#2333)

fs-eire committed 6 years ago

Verified da3c0ba1

Add FastGelu Cuda Op for Gelu and Add bias fusion (#2293)

tianleiwu committed 6 years ago

Verified b539cc74

Layer Normalization Fusion (#2319)

liuziyue committed 6 years ago

Verified 259bff8c

Add CUDA GatherElements kernel (#2310)

hariharans29 committed 6 years ago

Verified 553537ed

Make elementwise op run 4 items per thread (#2335)

yufenglee committed 6 years ago

Verified 6651d2f6

update dockerfiles/README (#2336)

jywu-msft committed 6 years ago

Verified ba0e7daf

[NupharEP] tensorize int8 GEMM for avx (#2142)

baowenlei committed 6 years ago

0f1e24f4

Newer Older