AgeCommit message (Collapse)AuthorFilesLines
2019-10-17[m2/neurun] Introduce build option for default config (#8206)tizen_5.5.m2_releasesubmit/tizen_5.5/20191031.000003submit/tizen/20191017.022054accepted/tizen/unified/20191017.105749accepted/tizen/5.5/unified/20191031.024450tizen_5.5accepted/tizen_5.5_unified오형석/On-Device Lab(SR)/Staff Engineer/삼성전자6-1/+19
- Introduce ENVVAR_NEURUN_CONFIG option: use default neurun config using environment (default: on) - Default off on tizen, and introduce gbs build option "test_build" to enable environment variable setting Change-Id: Ief8be67a56c0128b709d52580b9a2720b98bd428 Signed-off-by: Hyeongseok Oh <> Signed-off-by: Chunseok Lee <>
2019-10-15[m2/tool] Validation model file check in tflite_benchmark (#8150)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자1-1/+20
Pass verifier and use VerifyAndBuildFromFile in tflite_benchmark Signed-off-by: Hyeongseok Oh <>
2019-10-15[m2] Model file verification (#8134)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자1-3/+10
Verification check before tflite model loading Signed-off-by: Hyeongseok Oh <>
2019-09-26Do not install libraries with devel packages. (#7776)이춘석/On-Device Lab(SR)/Staff Engineer/삼성전자1-4/+0
Do not install duplicated libraries with devel packages; it incur build errors in applications with indirect dependencies on nnfw. For example, if A Requires B and B BuildRequires nnfw, building A indirect-BuildRequires nnfw's *.so files, which cannot be resolved by GBS or OBS because both nnfw and nnfw-devel have nnfw's *.so files. Thus, if nnfw has all *.so files in _libdir, do not install them with devel except for representative symlinks. Signed-off-by: MyungJoo Ham <>
2019-09-26Fix nnfw.h application build errors (#7775)이춘석/On-Device Lab(SR)/Staff Engineer/삼성전자1-0/+1
Fix build errors of applications using nnfw.h: ``` [ 15s] /usr/include/nnfw.h:162:35: error: unknown type name 'nnfw_tensorinfo' [ 15s] nnfw_tensorinfo *tensor_info); [ 15s] ^~~~~~~~~~~~~~~ [ 15s] /usr/include/nnfw.h:174:36: error: unknown type name 'nnfw_tensorinfo' [ 15s] nnfw_tensorinfo *tensor_info); [ 15s] ^~~~~~~~~~~~~~~ ``` Signed-off-by: MyungJoo Ham <>
2019-09-18[neurun] check null before accessing session (#7536)이상규/On-Device Lab(SR)/Principal Engineer/삼성전자1-2/+27
session will be null-checked before dereferencing. Signed-off-by: Sanggyu Lee <>
2019-09-18[res] Unit test of RealDiv 001 (#7549)박세희/On-Device Lab(SR)/Principal Engineer/삼성전자2-0/+64
This will add RealDiv unit test 001 for broadcasting Signed-off-by: SaeHie Park <>
2019-09-18[moco-tf] Add ResolveSquaredDifference knob (#7558)박세희/On-Device Lab(SR)/Principal Engineer/삼성전자1-0/+1
This will add ResolveSquaredDifference knob Signed-off-by: SaeHie Park <>
2019-09-18Make to support Neg op for acl neon (#7535)장지섭/On-Device Lab(SR)/Engineer/삼성전자11-1/+432
This commit makes to support Neg op for acl neon. - Introduce NEElementwiseUnaryKernel and NENegLayer - Apply NENegLayer layer for neurun Signed-off-by: jiseob.jang <>
2019-09-18[moco-tf] Use tf as prefix (#7489)박종현/On-Device Lab(SR)/Staff Engineer/삼성전자1-1/+1
Let's use "tf" prefix instead of "TF" to be compatiable with canonical prefix. Signed-off-by: Jonghyun Park <>
2019-09-17[locop] Introduce GenericNodeSummaryBuilder (#7518)박종현/On-Device Lab(SR)/Staff Engineer/삼성전자3-0/+141
This commit introduces GenericNodeSummaryBuilder which uses "dialect" address and opnum as opname. Signed-off-by: Jonghyun Park <>
2019-09-17[loco] Backward plane inference (#7458)박천교/On-Device Lab(SR)/Engineer/삼성전자1-0/+44
This commit introduces backward plane inference, which would be used for canonical shape inference rule, especially for TransposedConv2D. Signed-off-by: Cheongyo Bahk <>
2019-09-17[nnc] Remove C generating backend (#7545)Сергей Баранников/AI Tools Lab /SRR/Engineer/삼성전자11-355/+159
Remove C generating backend skeleton. Signed-off-by: Sergei Barannikov <>
2019-09-17Make to support Dequantize op for acl neon (#7522)장지섭/On-Device Lab(SR)/Engineer/삼성전자4-0/+21
This commit makes to support Dequantize op for acl neon. Signed-off-by: jiseob.jang <>
2019-09-17Make to support ReduceSum op for acl neon (#7520)장지섭/On-Device Lab(SR)/Engineer/삼성전자8-1/+295
This commit makes to support ReduceSum op for acl neon Signed-off-by: jiseob.jang <>
2019-09-17[exo-tflite] Introducing locoex::TFLConst (#7519)윤현식/On-Device Lab(SR)/Principal Engineer/삼성전자4-1/+76
locoex::TFLConst is added to represent tensor data. (note that this is not for a specific tflite op) Signed-off-by: Hyun Sik Yoon <>
2019-09-17[enco] Extract Boilerplate as Helper (#7523)박종현/On-Device Lab(SR)/Staff Engineer/삼성전자1-156/+120
This commit extracts boilerplate in ConstantFoldingTest as a helper in order to reduce code duplication. Signed-off-by: Jonghyun Park <>
2019-09-17Null-terminating check for backend conf api (#7516)이춘석/On-Device Lab(SR)/Staff Engineer/삼성전자2-5/+34
* Null-terminating check for backend conf api - Add null terminaing function - Add check routine Signed-off-by: Chunseok Lee <> * Give name to arg max length in backend conf - Add macro for max length - 32 for backend name, 64 for op name * fix format * add null check * add null check to nnfw_set_op_backend * Check null inside nnfw_session
2019-09-17[moco-tf] get_node_shape helper in copy_shapedata (#7528)박세희/On-Device Lab(SR)/Principal Engineer/삼성전자1-0/+16
This will introduce get_node_shape helper in copy_shapedata of FixShape Signed-off-by: Jonghyun Park <> Signed-off-by: SaeHie Park <>
2019-09-17[moco-tf] copy shape for binary input nodes (#7525)박세희/On-Device Lab(SR)/Principal Engineer/삼성전자1-12/+6
This will update FixShape to use correct copy_shapedata() for binary input nodes Signed-off-by: SaeHie Park <>
2019-09-17Remove "moco" module (#7509)박종현/On-Device Lab(SR)/Staff Engineer/삼성전자3-15/+0
No compiler module depends on "moco" module. Signed-off-by: Jonghyun Park <>
2019-09-17[plier-tf] Declare dependent library visibility (#7511)박종현/On-Device Lab(SR)/Staff Engineer/삼성전자1-2/+3
Let's declare the visibility of dependent libraries explicitly. This change allows us to introduce PRIVATE dependencies. Signed-off-by: Jonghyun Park <>
2019-09-17[plier-tf] Remove unnecessary const qualifier (#7513)박종현/On-Device Lab(SR)/Staff Engineer/삼성전자2-4/+4
"const" qualifier on function return type has no effect. Signed-off-by: Jonghyun Park <>
2019-09-17[logo] Simplifying `BiasEncode - BiasDecode` subgraph (#7481)윤현식/On-Device Lab(SR)/Principal Engineer/삼성전자1-0/+12
* [logo] Simplifying `BiasEncode - BiasDecode` subgraph With this commit, logo will Simplifying `BiasEncode - BiasDecode` subgraph into loco::Forward Signed-off-by: Hyun Sik Yoon <> * make code shorter * remove 'using namespace loco;'
2019-09-17Make to support ArgMax op for acl neon (#7515)장지섭/On-Device Lab(SR)/Engineer/삼성전자8-1/+243
This commit makes to support ArgMax op for acl neon except int32 type. - Introduce NEArgMinMax layer - Apply NEArgMinMax layer for neurun Signed-off-by: jiseob.jang <>
2019-09-17[tflite2circle] Introduce tflite2circle (#7483)이상규/On-Device Lab(SR)/Principal Engineer/삼성전자3-0/+129
It introduces `tflite2circle` tool. Its input is `tflite` and output is `circle`. Signed-off-by: Sanggyu Lee <>
2019-09-17[moco-tf] copy_shapedata with two inputs (#7499)박세희/On-Device Lab(SR)/Principal Engineer/삼성전자1-0/+58
* [moco-tf] copy_shapedata with two inputs This will introduce copy_shapedata that can receive two inputs and return shape from broadcast algorithm for binary operation nodes Signed-off-by: SaeHie Park <> * add note
2019-09-17[exo-tflite] Enhancing TestGraph.h (#7469)윤현식/On-Device Lab(SR)/Principal Engineer/삼성전자1-4/+31
* [exo-tflite] Enhancing TestGraph.h 1) Adds more operations. 2) changed the symantics of append() with no arg. Signed-off-by: Hyun Sik Yoon <> * typo, adds better comment
2019-09-17[exo-tflite] Introduce Div and Sub (#7514)박세희/On-Device Lab(SR)/Principal Engineer/삼성전자4-8/+66
This will introduce IR for Div and Sub Signed-off-by: SaeHie Park <>
2019-09-17Add ness dir into tizen test pacakge (#7478)이춘석/On-Device Lab(SR)/Staff Engineer/삼성전자1-2/+6
- Add infra dir - Fix tests/framework install position Signed-off-by: Chunseok Lee <>
2019-09-17[loco] Introduce GraphOutputIndexQueryService (#7479)박종현/On-Device Lab(SR)/Staff Engineer/삼성전자4-3/+71
This commit introduces GraphOutputIndexQueryService interface, and revises output_nodes helper to use this new service interface. This commit also implements GraphOutputIndexQueryService for canonical dialect in order to guarantee backward compatibility. Signed-off-by: Jonghyun Park <>
2019-09-17[loco] Extend CanonicalShapeInferenceRule with v2 API (#7468)박종현/On-Device Lab(SR)/Staff Engineer/삼성전자3-3/+113
This commit extends CanonicalShapeinferenceRule with v2 API implementation. Signed-off-by: Jonghyun Park <>
2019-09-17[moco-tf] Read Knob from top-level Canonicalizer (#7495)박종현/On-Device Lab(SR)/Staff Engineer/삼성전자6-21/+6
Several canonicalization implementations shows different behavior according to Knob configuration. This design makes it a bit difficult to test each implementation in isolation. To address this issue, this commit extracts Knob control code and put them in top-level Canonicalizer. Signed-off-by: Jonghyun Park <>
2019-09-17[locop] Show data type if possible (#7492)박종현/On-Device Lab(SR)/Staff Engineer/삼성전자1-2/+42
* [locop] Show data type if possible LinearV1 Graph Formatter now shows the data type of each node. Signed-off-by: Jonghyun Park <> * Fix a typo
2019-09-17[exo-tflite] Renaming get_knob -> get (#7510)윤현식/On-Device Lab(SR)/Principal Engineer/삼성전자4-10/+10
`get_knob` was renamed to `get` since it always use with knob name and `knob` is duplicate. (e.g., `get<Knob::EnableTFLDialect>`) Signed-off-by: Hyun Sik Yoon <>
2019-09-17Add neurun test scripts for tizen (#7504)이춘석/On-Device Lab(SR)/Staff Engineer/삼성전자2-0/+68
- neurun test script for acl_cl and mixed Signed-off-by: Chunseok Lee <>
2019-09-17Run neurun test on tizen (#7505)이춘석/On-Device Lab(SR)/Staff Engineer/삼성전자1-0/+6
- Run new neurun test script on tizen - Remove unnecessary zipped file Signed-off-by: Chunseok Lee <>
2019-09-17Update nnapi (#7482)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자2-99/+61
Use nnapi in android-10 r2 release Signed-off-by: Hyeongseok Oh <>
2019-09-17[nnpkg] rename schema name and update specification (#7484)이상규/On-Device Lab(SR)/Principal Engineer/삼성전자2-1/+5
nnpackage can have two kinds of models. `nnpackge_model.fbs` is for `circle`. Signed-off-by: Sanggyu Lee <>
2019-09-17Use string as backend configuration arg (#7500)이춘석/On-Device Lab(SR)/Staff Engineer/삼성전자4-90/+20
* Use string as backend configuration arg - remove unness. enums for backend/operation - Use string as backend api arguments Signed-off-by: Chunseok Lee <> * Fix backend conf api - Remove unness. functions - Reflect new backend conf api Signed-off-by: Chunseok Lee <> * Use const char* in c++ header
2019-09-17Update tizen's skiplist (#7475)이춘석/On-Device Lab(SR)/Staff Engineer/삼성전자1-49/+14
* Update tizen's skiplist Add arm-linux's skiplist into tizen's Signed-off-by: Chunseok Lee <> * Use arm-ubuntu's skiplist on tizen - Now, tizen's skiplist is the same as ubuntu's
2019-09-17[locomotiv] Simplify ReLU6 implementation (#7506)박종현/On-Device Lab(SR)/Staff Engineer/삼성전자1-0/+15
Let's make ReLU6 similar to other element-wise unary operations. Signed-off-by: Jonghyun Park <>
2019-09-17[loco] Type inference for TransposedConv2D (#7508)박천교/On-Device Lab(SR)/Engineer/삼성전자1-0/+1
This commit introduces type inference for TransposedConv2D Signed-off-by: Cheongyo Bahk <>
2019-09-17[mocotest-tf] Add regular testcases (#7488)박종현/On-Device Lab(SR)/Staff Engineer/삼성전자1-0/+9
This commit registers several testcases as mocotest-tf. Signed-off-by: Jonghyun Park <>
2019-09-17[loco] TransposedConv2D test case (#7498)박천교/On-Device Lab(SR)/Engineer/삼성전자1-0/+40
This commit adds test case for TransposedConv2D in loco Service Signed-off-by: Cheongyo Bahk <>
2019-09-17[exo-tflite] Modifying logging of each TFL nodes (#7507)윤현식/On-Device Lab(SR)/Principal Engineer/삼성전자1-11/+25
1) "" -> "", 2) op name printing in one method (opname()) Signed-off-by: Hyun Sik Yoon <>
2019-09-17[locomotiv] Erase annotated data before execute (#7496)박종현/On-Device Lab(SR)/Staff Engineer/삼성전자21-23/+2
All the execute method implementations invoke "erase_annot_data" internally. Signed-off-by: Jonghyun Park <>
2019-09-17[exo-tflite] Comment for TensorBroadcast (#7473)박세희/On-Device Lab(SR)/Principal Engineer/삼성전자1-0/+2
* [exo-tflite] Comment for TensorBroadcast This will add comment for TensorBroadcast in CanonicalNodeConverter template list as an ite Signed-off-by: SaeHie Park <> * edit message * fix typo and sentence
2019-09-17[exo-tflite] Resolve TensorBroadcast for Mul (#7491)박세희/On-Device Lab(SR)/Principal Engineer/삼성전자1-2/+13
This will add handler for TFLMul node in TensorBroadcastConverter Signed-off-by: SaeHie Park <>
2019-09-17[exo-tflite] adding FeatureBiasAdd converter (#7467)윤현식/On-Device Lab(SR)/Principal Engineer/삼성전자3-1/+117
* [exo-tflite] adding FeatureBiasAdd converter This adds FeatureBiasAdd converter, which converts loco::FeatureBiasEEdd to locoex::TFLAdd. Signed-off-by: Hyun Sik Yoon <> * typo