microsoft/onnxruntime

Pull Requests Commits

Minor optimization: if a node has already been placed, there's no need to find a kernel for it.

Pranav Sharma committed 6 years ago

abe40fb3

[NupharEP] force some low/zero cost ops to be inlined (#2409)

KeDengMS committed 6 years ago

Verified 1e03ce84

Add opset 11 versions of the existing CUDA operators that had negative axis support explicitly added. (#2398)

skottmckay committed 6 years ago

Verified c1d757a0

Change CUDA implementation of Transpose to support all fixed size tensor types (#2387)

skottmckay committed 6 years ago

Verified 6cc57721

Avoid using the default logger in the graph lib and optimizers (#2361)

Changming Sun committed 6 years ago

Verified 109b3cb4

[NupharEP] Multiple optimizations (#2380)

KeDengMS committed 6 years ago

Verified b15e43a5

Fix reuse logic in allocation planner. (#2393)

Pranav Sharma committed 6 years ago

Verified 7e164eaa

Fixed compilation with ngraph (#2388)

ilya-lavrenov committed 6 years ago

b90d55b7

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

Older