microsoft/onnxruntime

Pull Requests Commits

Disable DML EP on software adapter, fix float16 fallback bug, re-enable DML in CI (#2896)

jeffbloo committed 5 years ago

Verified ba336b55

Fix some warnings on Windows (#2560)

Changming Sun committed 5 years ago

Verified 201b089a

Disable openmp for the nocontribops pipeline. (#2888)

Pranav Sharma committed 5 years ago

Verified 49725f89

Update BFCArena logic to use backoff if cudaMalloc fails. Makes behaviour equivalent to when a CPU allocation fails. Add unit test. (#2748)

skottmckay committed 5 years ago

Verified fc51473b

Fixed typo in ORT_RETURN_IF_NOT() message. (#2862)

edgchen1 committed 5 years ago

061f10fc

InferenceSession::Run needs to call OnRunEnd for any EP that OnRunStart was called for so they can cleanup. Currently it only calls OnRunEnd if the Status is OK. Due to this the CUDA EP will throw during shutdown as the per-thread information has not been cleaned up prior to the CUDA library shutting down. (#2881)

skottmckay committed 5 years ago

Verified 9f5e8c4a

Rashuai/boost cuda TopK performance (#2826)

RandySheriffH committed 5 years ago

Verified 38b34bab

Optimize BatchNormalization to NCHWc Conv (#2855)

tracysh committed 5 years ago

Verified 08113b80

Add calibration tool (#2845)

askhade committed 5 years ago

Verified 807a59c5

Fix positive raw scores for TreeEnsembleClassifier (#2824)

xadupre committed 5 years ago

Verified 22d9f399

Support non-sequence tensor fed through as a python list (#2782)

hariharans29 committed 5 years ago

b21576ee

Fix spurious component detection warning (#2857)

KeDengMS committed 5 years ago

Verified f9f25ec0

Add float16 support back in the bert fusion script (#2870)

yufenglee committed 5 years ago

Verified 25d7ad18

Bert fusion script for Tensorflow squad (#2858)

yufenglee committed 5 years ago

Verified 95f3eb6a

update protoc path to match protobuf version (#2865)

tracysh committed 5 years ago

Verified 01f3a33c

Update Windows GPU build to use cudnn 7.6

Changming Sun committed 5 years ago

e6f7658a

Fix topk type handling to accommodate more types. (#2842)

Pranav Sharma committed 5 years ago

Verified 3853ddf9

Disable DML in Windows GPU CI build (#2856)

Changming Sun committed 5 years ago

Verified 47e27ec9

Use IArenaAllocator::Reserve for initializers and mem pattern planner blocks (#2835)

skottmckay committed 5 years ago

Verified 724ff075

MLAS: enable threading for quantized GEMMs (#2844)

tracysh committed 5 years ago

Verified 928b6bb2

update optimization doc for BERT related fusions (#2819)

tianleiwu committed 5 years ago

Verified 5db85430

Fix Linux CUDA nuget packaging pipeline break

Changming Sun committed 5 years ago

56030f8d

Fix cgmanifest.json generating script (#2770)

tiagoshibata committed 5 years ago

cff266e1

User/orilevari/32bit comparison warning (#2800)

Ori Levari committed 5 years ago

db05436f

add domain check for nodes + update documentation (#2831)

askhade committed 5 years ago

Verified 8643f3eb

Convert ExternalProject Featurizers into git submodule (#2834)

yuslepukhin committed 5 years ago

Verified aa37dea5

Ignore allocator type in ExecutionProviders allocator map. Make default initialization of OrtMemoryInfo more clearly invalid. (#2768)

skottmckay committed 5 years ago

Verified 98cb41aa

Add support for int64_t for topk CPU. Fixes github issue #2806. (#2833)

Pranav Sharma committed 5 years ago

Verified b308e826

Upgrade gtest to the latest version (#2827)

Changming Sun committed 5 years ago

Verified 5c391854

Add OneHotEncoder and HashOneHotEncoder kernels. (#2830)

yuslepukhin committed 5 years ago

Verified 120433c2

Newer Older