summaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2019-07-22Update TFLite version to v1.13.1submit/tizen/20190722.095424accepted/tizen/unified/20190723.112126tizen_5.5_m1_bakChunseok Lee5-12/+12
- Reflect recent update of tensorflow-lite-devel package Change-Id: I06a6c199ff4e2270280139fef25276c84ec4f356 Signed-off-by: Chunseok Lee <chunseok.lee@samsung.com>
2019-04-29Enable custom oprations test (#5072)submit/tizen/20190429.094136accepted/tizen/unified/20190430.113441오형석/On-Device Lab(SR)/Staff Engineer/삼성전자4-4/+0
Enable custom operations framework test again Change-Id: I02cf317122ac3d304107738ffc6abd17e665dcfa Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-23Manage operand stack and operation stack for interpretation (#5020)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자1-1/+59
- Introduce operand ready checker - Introduce operation stack to save prepared operation - Manage operand stack and operation stack and show log Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-23Introduce operand usage vectorin frontend (#5039)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자2-1/+26
Collect operand usage info in frontend Clear when model building is finished Prepare removing nnapi-dependent usage info in neurun model operand Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-23Introduce build option for tflite accuracy tool (#5048)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자2-0/+5
Introduce build option for tflite accuracy tool and set off as default Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-23Change default runtime to neurun (#5042)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자43-131/+107
* Change default runtime to neurun Change default runtime: pacl -> neurun Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com> * Enable loggin runtime default
2019-04-23[batch_run] A batch execution tool used to run experiments (#4507)Saulo Aldighieri Moraes/AI R&D /SRBR/Engineer/삼성전자4-0/+1563
* Related: #4405 The batch execution tool can be used to run experiments. It reads a neural network model from a file and a series of input images from a directory, runs each image through the network, and collect statistics, such as execution time and accuracy. Signed-off-by: Saulo A. Moraes <s.moraes@samsung.com> * Remove boost_ext reference Build script cleanup, remove unnecessary reference to boost_ext. Signed-off-by: Saulo A. Moraes <s.moraes@samsung.com> * Related: #4405 Code review fixes and tool rename from batch_run to tflite_accuracy. Signed-off-by: Saulo A. Moraes <s.moraes@samsung.com>
2019-04-23Remove operand usage getter methods (#5038)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자10-17/+11
- Remove unused isModelInput() method - Remove usage() method. Use hasData() and model's input/output Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-22Ignore try-run result on cross build (#5031)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자4-0/+11
* Ignore try-run result Ignore try-run result on cross build when finding cmake threads package Threads package is checked when finding pre-installed tflite package Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com> * Fix comment
2019-04-22Change ARGMAX definition (#5021)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자13-18/+35
Chagne ARGMAX definition same with current NNAPI - Allow one axis reduce - Output rank = input rank - 1 Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-19Add printing info for filled memory in model_parser (#5028)김용섭/On-Device Lab(SR)/Engineer/삼성전자1-2/+5
* Add printing info for filled memory in model_parser Add printing info for filled memory such as weight in model_parser Signed-off-by: Yongseop Kim <yons.kim@samsung.com> * fix failed format-check
2019-04-19Use POSITION_INDEPENDENT_CODE property for tflite build (#5026)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자1-1/+1
Use cmake POSITION_INDEPENDENT_CODE property instead of setting gcc -fPIC compile option directly Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-18[neurun] Apply code for merging node into subgraph to Graph (#5019)김용섭/On-Device Lab(SR)/Engineer/삼성전자1-22/+168
Apply code for merging node into subgraph to Graph instead of 1 subrgaph 1 node code Signed-off-by: Yongseop Kim <yons.kim@samsung.com>
2019-04-18Remove OBS_BUILD cmake define in Makefile (#5017)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자1-1/+0
Remove OBS_BUILD cmake define in cmake configure because OBS_BUILD is not used in cmake any more Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-18Prepare operand stack in interpreter (#5015)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자1-0/+29
Prepare operand stack in interpreter to save prepared operands to use Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-18[neurun] Apply SubgraphContext(1 subgraph 1 node) (#5014)김용섭/On-Device Lab(SR)/Engineer/삼성전자9-140/+207
Apply SubgraphContext(1 subgraph 1 node) to Graph, Linear, Compiler and {Dataflow|Linear}Executor Signed-off-by: Yongseop Kim <yons.kim@samsung.com>
2019-04-18Fix CMake to use STATUS for message (#5016)박세희/On-Device Lab(SR)/Principal Engineer/삼성전자1-6/+6
This will fix message() to use STATUS instead if "--" string Signed-off-by: SaeHie Park <saehie.park@samsung.com>
2019-04-17Interpreter ITensor Inherit backend ITensor (#5005)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자1-2/+2
Interpreter ITensor Inherit backend ITensor for more compatibility with other backends Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-17[neurun] Rename operation::Node as Operation (#5013)이한종/On-Device Lab(SR)/Engineer/삼성전자135-275/+276
This is for reducing namespace operation. Signed-off-by: Hanjoung Lee <hanjoung.lee@samsung.com>
2019-04-17Fix a couple of noticeable clang static analyzer warnings (#4981)Aleksei Grebenkin/AI Ecosystem Lab /SRR/Engineer/삼성전자3-6/+6
* Fix a couple of noticeable clang static analyzer warnings Fixed passing vector by value, lack of default switch case, dead code Signed-off-by: Alexey Grebenkin <a.grebenkin@samsung.com> * Correcting for PR comments * Correcting for more PR comments
2019-04-17Fix define in cker header file (#5007)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자2-6/+6
Fix define __NNFW_CKER_CONCATENATION_H_ => __NNFW_CKER_CONCATENATION_H__ Fix define __NNFW_CKER_FXIED_POINT_H__ => __NNFW_CKER_GEMMLOWP_FIXED_POINT_H__ Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-17Fix OperationIndex build break from Subgraph (#5012)이상규/On-Device Lab(SR)/Principal Engineer/삼성전자2-28/+40
It fixes build-break caused by Subgraph's use of model::operation::Index, which is renamed to model::OperationIndex. Signed-off-by: Sanggyu Lee <sg5.lee@samsung.com>
2019-04-17Introduce interpreter class (#5006)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자3-2/+117
* Introduce interpreter class Introduce interpreter class: run on assigned ExecEnv Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com> * Fix define for header
2019-04-16[neurun] Introduce SubgraphContext class (#4995)김용섭/On-Device Lab(SR)/Engineer/삼성전자2-0/+195
This class would be used as a context class which has Subgraphs instead of SubgraphSet in Graph Signed-off-by: Yongseop Kim <yons.kim@samsung.com>
2019-04-16Modify cpu average pool kernel (#5004)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자1-22/+26
Move loop invariant codes Calculate filter count at once and compare Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-16[neurun] Rename operation::Index as OperationIndex (#5002)이한종/On-Device Lab(SR)/Engineer/삼성전자30-105/+102
This is for reducing namespace operation. Part of #4985 Signed-off-by: Hanjoung Lee <hanjoung.lee@samsung.com>
2019-04-15Remove axis conversion in libs/ARMComputeEx (#4992)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자7-199/+30
Remove axis conversion in libs/ARMComputeEx because conversion should done before configuration Fix axis conversion in pack/unpack acl-cl extend kernel Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-15Move gemmlowp code in cker (#4998)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자3-12/+16
Move gemmlowp code in cker into cker/gemmlowp Add namespace for gemmlowp Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-15[neurun] Change from SubgraphSet to SubgraphSequence (#4996)김용섭/On-Device Lab(SR)/Engineer/삼성전자8-25/+28
Change the name SubgraphSet. Actually SubgraphSet has elements in order so that SubgraphSequence is proper name. Signed-off-by: Yongseop Kim <yons.kim@samsung.com>
2019-04-15Remove OpenCLEx codes (#4993)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자2-219/+0
- CLKernelLibraryEx includes OpenCL.h, not OpenCLEx.h - There is no problem now Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-15Interpreter use info structure in model (#4988)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자4-124/+51
Instead of using interpreter's own structure, use tensor info structure in model Move comment and copy constructor Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-15Remove script call in acl test (#4991)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자1-3/+0
Remove neurun acl-neon test script call in acl-cl test script It will be called directly on CI infra. Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-11Remove install backend kernel (#4979)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자3-3/+0
Remove install static kernel library because they are linked at each backend Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-11Move backend cmake setting into each backend (#4976)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자7-22/+21
Move backend cmake variable setting into each backend Remove target check in backend Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-11Add more coverage test (#4977)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자1-21/+4
* Add more coverage test Call scripts for each backend test Add acl_neon backend kernel test Add dataflow executor test Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com> * Fix script name typo
2019-04-11[neurun] add ExecManager::{setInput,setOutput} unit tests (#4972)이상규/On-Device Lab(SR)/Principal Engineer/삼성전자1-0/+94
Add unittest for specifying unspecified dimension with setInput and setOutput. Signed-off-by: Sanggyu Lee <sg5.lee@samsung.com>
2019-04-11Remove cpu backend build dependency with tflite (#4975)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자2-12/+0
Remove header path and linking dependency with tflite in cpu backend build cmake Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-11[ubun] Introduce Softmax microbenchmark (#4974)박종현/On-Device Lab(SR)/Staff Engineer/삼성전자2-0/+59
With this commit, ubun now includes a Softmax microbenchmark based on cker library. Signed-off-by: Jonghyun Park <jh1302.park@samsung.com>
2019-04-10[neurun] Remove type and shape argument from Executor's setInput and ↵이상규/On-Device Lab(SR)/Principal Engineer/삼성전자6-31/+132
setOutput (#4957) IExecutor, ExecutorBase, ExecManager will have setInput and setOutput that don't require `type` and `shape` parameters. In this case, `type` and `shape` from model will be used. ExecManager test is also modified to use newly introduced setInput and setOutput. Signed-off-by: Sanggyu Lee <sg5.lee@samsung.com>
2019-04-10Remove unused util function in cpu backend (#4971)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자1-73/+0
Remove unused util function in cpu backend: type conversion to tflite internal type Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-10[neurun] Framework test with acl_neon backend (#4961)이한종/On-Device Lab(SR)/Engineer/삼성전자4-0/+138
Enable framework test with acl_neon backend. This is a workaround since adding another job in CI has a lot more to do. Signed-off-by: Hanjoung Lee <hanjoung.lee@samsung.com>
2019-04-10Remove tflite macro and type in cker library (#4966)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자2-6/+6
Remove tflite macro TFLITE_DCHECK_XX in cker library Use uint8_t instead of tlfite type uint8 Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-10Introduce cpu quant8 softmax kernel (#4953)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자4-7/+340
Introduce cpu quantized int8 softmax kernel from tflite and gemmlowp Use kernel in neurun cpu backend Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-10Fix copyright (#4967)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자3-1/+33
Fix copyright using introduced script Add copyright in libs/ARMComputeEx/src/runtime/misc/functions/Utils.cpp, runtimes/neurun/backend/acl_neon/operand/INETensor.cc Fix libs/ARMComputeEx/arm_compute/runtime/misc/functions/Utils.h year based on file created date Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-10[NNFW] optimized Conv on CPU (#4964)Андрей Шедько/AI Tools Lab /SRR/Engineer/삼성전자1-20/+20
Moved condition and offset calculation out of the innermost loop in cpu convolution kernel. Achieved 2+x speed up on mobilenet on neurun. Signed-off-by: Andrei Shedko <a.shedko@samsung.com>
2019-04-10Move rootfs setting into cmake/buildtool/cross (#4958)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자10-69/+72
Move rootfs setting and check correctness into cmake/buildtool/cross Remove different variable name for each architecture rootfs Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-09Update HowToUseDockerImage.md (#4445)Sergei Chicherin/Vision, Learning and Telepresence Lab /SRR/Staff Engineer/삼성전자1-1/+1
mistype : missed / before etc Signed-off-by: Sergei Chicherin <s.chicherin@samsung.com>
2019-04-09Fix copyright checker (#4954)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자1-0/+9
Rename copyright check script Change mode: remove execution and allow only source (by nnfw command) Exit with exit code when invalid result Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-09Ignore directories generated by coverage scripts (#4960)오형석/On-Device Lab(SR)/Staff Engineer/삼성전자1-0/+4
Ignore directories generated by coverage scripts - coverage test run: /gcov - coverage analysis result: /coverage Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
2019-04-09[neurun] Remove redundant virtual from override functions (#4956)이상규/On-Device Lab(SR)/Principal Engineer/삼성전자14-52/+48
I removed `virtual` when I am sure it is our code. Otherwise, I did not removed `virtual`. Signed-off-by: Sanggyu Lee <sg5.lee@samsung.com>