오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 8 Nov 2018 08:15:49 +0000 (17:15 +0900)]
Fix typo in nnapi delegate (#3527)
Fix typo in nnapi delegate: v1.11 -> v1.12
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
김용섭/동작제어Lab(SR)/Engineer/삼성전자 [Thu, 8 Nov 2018 07:48:44 +0000 (16:48 +0900)]
[neurun] Apply IMemoryPlanner to backend/cpu (#3517)
Now IMemoryPlanner and Allocator plays the role of IMemoryAllocator.
This commits includes some fixing for bugs to apply on backend/cpu.
Signed-off-by: Yongseop Kim <yons.kim@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 8 Nov 2018 07:12:55 +0000 (16:12 +0900)]
Update tensorflow to v1.12 (#3486)
* Update external package
- Update external package version
- Update cmake to build tflite
- Add gitignore for new external code
* Update library to support tflite
- Update nnapi delegate
- Update NeuralNetworksShim.h for new API
- Fix cmake for new files in updated tensorflow
- Workaround: Introduce OBS_BUILD flag to avoid gbs build fail
* Update neurun CPU kernel
- Update neurun CPU kernel to call updated tflite interpreter kernel
- Add build option -fPIC for updated tflite kernel
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 8 Nov 2018 06:49:08 +0000 (15:49 +0900)]
Register subtensor info to tensor builder (#3519)
Register subtensor info to tensor builder if
- Tensor is subtensor of other tensor
- Backend support subtensor allocation
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 8 Nov 2018 06:40:53 +0000 (15:40 +0900)]
Check external directory exists (#3522)
Check external directory exists and download if directory does not exist
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
이춘석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 8 Nov 2018 05:12:20 +0000 (14:12 +0900)]
Fix run_benchmark.sh typo(MODLES->MODELS) (#3523)
- Fix run_benchmark.sh typo( MODLES -> MODELS )
Signed-off-by: Chunseok Lee <chunseok.lee@samsung.com>
이춘석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 8 Nov 2018 04:08:02 +0000 (13:08 +0900)]
Fix CI benchmark failure (#3520)
CI benchmark job fails since it does not reflect model path
change. This commit sets the proper model path.
Signed-off-by: Chunseok Lee <chunseok.lee@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 8 Nov 2018 02:44:39 +0000 (11:44 +0900)]
Concat elimination in codegen (#3518)
Concat elimination in codegen: check output tensor is parent of inputs and do nothing if true
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 8 Nov 2018 01:56:03 +0000 (10:56 +0900)]
Build subtensor in acl TensorBuilder (#3513)
* Build subtensor in acl TensorBuilder
Build subtensor in acl TensorBuilder
Change to public function in Convert.cc
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
* Remove deprecated declaration
* Remove diff by space line
이한종/동작제어Lab(SR)/Engineer/삼성전자 [Thu, 8 Nov 2018 01:11:02 +0000 (10:11 +0900)]
[neurun] Remove InitializerGenerator dependency (#3508)
Remove all InitializerGenerator dependent code since we now have
ConstantInitializer for this purpose.
Signed-off-by: Hanjoung Lee <hanjoung.lee@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 8 Nov 2018 01:10:32 +0000 (10:10 +0900)]
Introduce subtensor allocation check API (#3510)
Check API will be used in codegen
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
김용섭/동작제어Lab(SR)/Engineer/삼성전자 [Thu, 8 Nov 2018 01:09:12 +0000 (10:09 +0900)]
[neurun] Introduce MemoryPlanner and Allocator (#3511)
Introduces IMemoryPlanner, BumpPlanner, FirstFitPlanner and
Allocator to replace IMemoryAllocator.
Current IMemoryAllocator has two roles, Planner and Allocator. This
commits separates IMemoryAllocator to the two classes. After this commit
lands, IMemoryAllocator will be removed soon.
Signed-off-by: Yongseop Kim <yons.kim@samsung.com>
장지섭/동작제어Lab(SR)/Engineer/삼성전자 [Wed, 7 Nov 2018 09:31:46 +0000 (18:31 +0900)]
Optimize the StridedSlice Kernel for performance improvement (#3230)
This commit optimizes the StridedSlice Kernel for performance inprovement.
Signed-off-by: jiseob.jang <jiseob.jang@samsung.com>
Introducing NotEqual Op in runtime (#3485)
This commit introduces NotEqual op in runtime.
Related issues: #3335, #3337
Signed-off-by: Siva Sai <siva.sai@samsung.com>
ArgMinMax Implementation for ArgMin and ArgMax. (#3331)
This change introduces the cpu kernel for the ArgMin and ArgMax
operations.
Signed-off-by: Praveen D N <praveen.dn@samsung.com>
이한종/동작제어Lab(SR)/Engineer/삼성전자 [Wed, 7 Nov 2018 08:01:26 +0000 (17:01 +0900)]
[neurun] Fix build break (#3507)
Caused by #3426 and #3436 collision.
Signed-off-by: Hanjoung Lee <hanjoung.lee@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Wed, 7 Nov 2018 07:40:33 +0000 (16:40 +0900)]
Update framework test list (#3499)
Add more framework test for neurun
Change inception test: use simple module instead of full models
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Wed, 7 Nov 2018 07:17:33 +0000 (16:17 +0900)]
Introduce subtensor map (#3493)
- Introduce subtensor map in ACL tensor builder
- TensorBuilder::at() and TensorBuilder::tensorAt() can return subtensor
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
이한종/동작제어Lab(SR)/Engineer/삼성전자 [Wed, 7 Nov 2018 07:02:25 +0000 (16:02 +0900)]
[neurun] Introduce ConstantInitializer (#3426)
* [neurun] Introduce ConstantInitializer
This commit introduces `ConstantIntializer` class that fills operand
objects with the constant values which were specified by
`ANeuralNetworksModel_setOperandValue`. Plus, this class replaces the
InitializerGenerator.
NOTE due to FullyConnected operation's implicit reshaping, it is
implemented with workarounds.
Signed-off-by: Hanjoung Lee <hanjoung.lee@samsung.com>
* Add license
Signed-off-by: Hanjoung Lee <hanjoung.lee@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Wed, 7 Nov 2018 06:34:34 +0000 (15:34 +0900)]
Fix custom operator register macro (#3495)
Fix custom operator register macro for updated TfLiteRegistration field
- First 4 fields may not change
- Can adjust new fields before custom_name
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
이춘석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Wed, 7 Nov 2018 06:34:20 +0000 (15:34 +0900)]
Enable more model tests (#3497)
The following tests are triggered on default on framework/verification
test :
- embedding lookup
- L2 normalization
- L2 pool
- Squeeze
- Tanh
- Transpose
- Logistic
Signed-off-by: Chunseok Lee <chunseok.lee@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Wed, 7 Nov 2018 06:18:40 +0000 (15:18 +0900)]
Introduce namespace in runtime load helper (#3494)
Introduce namespace in runtime load helper
Restore function name
It protects naming conflict with upstream tensorflow
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
김용섭/동작제어Lab(SR)/Engineer/삼성전자 [Wed, 7 Nov 2018 03:12:34 +0000 (12:12 +0900)]
[model_parser] Hide instruction counts by verbose option (#3487)
Hides instruction counts by verbose option. Now -v or --verbose can skip
printing instruction counts. You can use this verbose option as follows
- With verbose level 0, prints only summary of tflite
- With verbose level 1(default), prints all like current default except for
instructions
- With verbose level 2, prints all like current default including for
instructions
Signed-off-by: Yongseop Kim <yons.kim@samsung.com>
김용섭/동작제어Lab(SR)/Engineer/삼성전자 [Tue, 6 Nov 2018 09:15:59 +0000 (18:15 +0900)]
[neurun] Introduce FirstfitAllocator (#3458)
* [neurun] Introduce FirstfitAllocator
Introduces FirstfitAllocator instead of BumpAllocator. This allocator
allocates a block for tensor at the first-fit offset.
Signed-off-by: Yongseop Kim <yons.kim@samsung.com>
* Revise allocate() of FirstfitAllocator
- Adds comment for description of allocate()
- Removes inner lambda allocateBlock() in allocate()
* Change the name from Firstfit to FirstFit
* Fix typo
* Adds comment for _alloc_table
장지섭/동작제어Lab(SR)/Engineer/삼성전자 [Tue, 6 Nov 2018 08:50:30 +0000 (17:50 +0900)]
Introduce a enum class declaration for reduce operations. (#3461)
This commit introduces a enum declaration for reduce operations.
Signed-off-by: jiseob.jang <jiseob.jang@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Tue, 6 Nov 2018 02:48:23 +0000 (11:48 +0900)]
Introduce registerSubTensorInfo (#3474)
Introduce registerTensorInfo to register subtensor information (not used yet)
Backend allocator will use this info for allocation
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
Introducing NotEqual op in NeuralNetworksEx (#3473)
This commit introduces NotEqual op in NeuralNetworksEx.h
Related issues: #3335, #3337
Signed-off-by: Siva Sai <siva.sai@samsung.com>
Add CL kernel call for Transpose op from runtime (#3470)
This patch adds cl kernel call for transpose op from PACL.
Signed-off-by: shubham <shub98.gupta@samsung.com>
Updating comment to create more formal equation (#3466)
This patch will updates comment for sqrt so as to remove x and y and to
have more formal equation.
Signed-off-by: shubham <shub98.gupta@samsung.com>
Enable broadcast support for Equal_Ex op (#3431)
This patch enables broadcast support for Equal_Ex op.
Related issue: #3295.
Signed-off-by: prasannar <prasanna.r@samsung.com>
Correction of comments by adding shape info in NeuralNetworksEx.h (#3463)
This patch will add correction in comments by adding shape info in NeuralNetworksEx.h
Signed-off-by: shubham <shub98.gupta@samsung.com>
Defining CAST_EX in NeuralNetworksEx.h (#3424)
This patch defines CAST_EX in NeuralNetworksEx.h
Signed-off-by: shubham <shub98.gupta@samsung.com>
Defining RSQRT_EX in NeuralNetworksEx.h (#3415)
This patch defines RSQRT_EX in NeuralNetworksEx.h
Signed-off-by: shubham <shub98.gupta@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Mon, 5 Nov 2018 08:01:56 +0000 (17:01 +0900)]
Introduce registerTensorInfo (#3467)
Introduce registerTensorInfo to
- Divide tensor info register and use-def notify
- Prepare subtensor information register
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Mon, 5 Nov 2018 07:43:03 +0000 (16:43 +0900)]
Fix build failure on neuurun (#3468)
Fix build failure on neurun by operator structure change
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
[neurun] Structural changes for ops (#3436)
This commit follows the discussions on #3250
Signed-off-by: Siva Sai <siva.sai@samsung.com>
장지섭/동작제어Lab(SR)/Engineer/삼성전자 [Mon, 5 Nov 2018 07:04:39 +0000 (16:04 +0900)]
Support ReduceMax kernel for cl up to 4-dimensions (#3340)
* Support ReduceMax kernel for cl up to 4-dimensions
This commit supports ReduceMax kernel for cl up to 4-dimensions.
Signed-off-by: jiseob.jang <jiseob.jang@samsung.com>
* Optimize ReduceMax Kernel for cl
This commit optimizes ReduceNMax kernel for cl.
- Change calling kernel from at once kernel to call separated kernels multiple times.
Signed-off-by: jiseob.jang <jiseob.jang@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Mon, 5 Nov 2018 05:58:51 +0000 (14:58 +0900)]
Introduce backend config for subtensor (#3460)
Introduce backend config for subtensor
CPU and ACL backend are not support subtensor yet
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
Comment correction in CLBatchToSpaceND.h (#3313)
This patch will correct the doxygen comment for CLBatchToSpaceND.h
Signed-off-by: shubham <shub98.gupta@samsung.com>
김용섭/동작제어Lab(SR)/Engineer/삼성전자 [Mon, 5 Nov 2018 01:12:03 +0000 (10:12 +0900)]
[neurun] Apply tensor lifetime-info(use-def) to Linear/TensorBuilder (#3409)
* [neurun] Apply tensor lifetime-info(use-def) to memory allocation
Applies tensor lifetime-info(use-def) to memory allocation with revising
Linear and TensorBuilder.
- Introduce vassignTensors(virtually assign tensors) in Linear instead
of markTensors
- Introduce assign(assign memory virtually)/unassign(unassign memory
virtually) in TensorBuilder instead of mark
Signed-off-by: Yongseop Kim <yons.kim@samsung.com>
* Fix release build error
* Rename functions properly
Linear::vassignTensors -> planTensors
ITensorBuilder::assign -> notifyFirstUse
ITensorBuilder::unassign -> notifyLastUse
* Fix log
* Fix searching output first than input
Implementation of Generic Permute CL Kernel (#3114)
This patch will add generic permute kernel supporting all permutations for nD Tensors
This kernel is required for Reshape and Transpose op
Signed-off-by: shubham <shub98.gupta@samsung.com>
Added test cases for Pack operation (#3399)
Test cases for 3D to 4D pack along 2 axises
Added test in skiplist of arm-linux, neuran and x86-linux
Signed-off-by: Vishal Keshav <vishal.k1@samsung.com>
Defining NEG_EX in NeuralNetworksEx.h (#3417)
This patch defines NEG_EX in NeuralNetworksEx.h
Related issue: #3347
Signed-off-by: prasannar <prasanna.r@samsung.com>
Defining EXP_EX in NeuralNetworksEx.h (#3416)
This patch defines EXP_EX in NeuralNetworksEx.h
Related issue: #3347
Signed-off-by: prasannar <prasanna.r@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 1 Nov 2018 23:49:29 +0000 (08:49 +0900)]
Remove deprecated coverity script (#3440)
Remove deprecated coverity script (used nowhere)
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 1 Nov 2018 08:21:17 +0000 (17:21 +0900)]
Remove svace and coverity build script (#3439)
Remove svace and coverity build script
- These scripts are not used in CI any more
- These scripts are not used by developer
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
Remove unnecessary headers from src/core/CL/kernels/* (#3375)
This patch will remove unnecessary headers from src/core/CL/kernels/* files
Signed-off-by: shubham <shub98.gupta@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 1 Nov 2018 04:37:00 +0000 (13:37 +0900)]
Revise script for docker image build (#3430)
Revise script for docker image build
- Change script name
- Rename common.sh and move
- Set argument for docker image name and tizen build image
- Update documents
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 1 Nov 2018 04:36:46 +0000 (13:36 +0900)]
Introduce SubTensorAnalyzer (#3410)
Introduce SubTensorAnalyzer and insert subtensor analyze phase in compilation
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
이한종/동작제어Lab(SR)/Engineer/삼성전자 [Wed, 31 Oct 2018 10:58:29 +0000 (19:58 +0900)]
[neurun] Introduce iterate method for Context (#3425)
Introduce `iterate` method for `codegen::operand::Context`.
Signed-off-by: Hanjoung Lee <hanjoung.lee@samsung.com>
Move Graph::insertPermute to PermutationInsertionPass (#3408)
Since this method is used just in `PermutationInsertionPass` it's
reasonable to move into this class
Related issue is https://github.sec.samsung.net/STAR/nnfw/issues/3388
Signed-off-by: Poshshoev Dilshodzhon <d.poshshoev@samsung.com>
Defining SQUARED_DIFFERENCE_EX in NeuralNetworksEx.h (#3413)
This patch defines SQUARED_DIFFERENCE_EX in NeuralNetworksEx.h
Related issue: #3347
Signed-off-by: prasannar <prasanna.r@samsung.com>
Defining EQUAL_EX in NeuralNetworksEx.h (#3354)
This patch defines EQUAL_EX in NeuralNetworksEx.h
Related issue: #3347
Signed-off-by: prasannar <prasanna.r@samsung.com>
[neurun] Modification for Add op (#3414)
The changes made are aligned with the discussion in #3250
Signed-off-by: Siva Sai <siva.sai@samsung.com>
Introduce SQRT in PACL (#3412)
This patch introduces SQRT op in PACL
Related issue: #3341
Signed-off-by: prasannar <prasanna.r@samsung.com>
Introduce SQRT_EX in NeuralNetworksEx.h (#3400)
This patch includes SQRT_EX in NeuralNetworksEx.h
Related issue: #3341
Signed-off-by: prasannar <prasanna.r@samsung.com>
이상규/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 31 Oct 2018 05:50:48 +0000 (14:50 +0900)]
Run benchmark for all tflite model under tests/framework (#3402)
Related: #3397 #3264
Previously, run_benchmark_op.sh runs the tests in `benchmark_op_models.txt`.
- No one maintains the list.
- It contains only a few tests.
- If the location or folder name is changed, the list should be updated.
Now, run_benchmark_op.sh runs all tests under `tests/frameworks`.
- No one need to update whenever a new model is added/removed, location is changed....
This PR also removes
- option `--frameworktest_list_file` option from `run_benchmark_op.sh`.
- benchmark_op_list.txt
Signed-off-by: Sanggyu Lee <sg5.lee@samsung.com>
이한종/동작제어Lab(SR)/Engineer/삼성전자 [Wed, 31 Oct 2018 03:59:34 +0000 (12:59 +0900)]
[neurun] Introduce OperationPass (#3403)
Introduce `OperationPass` class which iterates over `operation::Node`s.
This is the same manner with `OperandPass`.
Signed-off-by: Hanjoung Lee <hanjoung.lee@samsung.com>
Removing TODO for pool op (#3370)
This patch will remove TODO statement for max_pool and avg_pool
Signed-off-by: shubham <shub98.gupta@samsung.com>
윤지영/동작제어Lab(SR)/Engineer/삼성전자 [Wed, 31 Oct 2018 01:22:27 +0000 (10:22 +0900)]
Use rpm macro instead of user define macro in spec file (#3406)
%{install_prefix} -> %{_prefix}
%{install_prefix}/lib -> %{_libdir}
Signed-off-by: Jiyoung Yun <jy910.yun@samsung.com>
Adding Unpack Node (#3216)
Added unpack node in pure_arm_compute
Signed-off-by: Vishal Keshav <vishal.k1@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Tue, 30 Oct 2018 12:06:07 +0000 (21:06 +0900)]
Introduce SubTensorInfo class (#3407)
Introduce SubTensorInfo class to use at backend tensor allocation
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Tue, 30 Oct 2018 07:58:47 +0000 (16:58 +0900)]
Introduce ParentInfo class (#3401)
Introduce ParentInfo class to represent parent tensor information
Add field in neurun::graph::operand::Object for ParentInfo
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
김용섭/동작제어Lab(SR)/Engineer/삼성전자 [Tue, 30 Oct 2018 06:07:18 +0000 (15:07 +0900)]
[neurun] Revise MemoryAllocator (#3386)
- MemoryBlock -> Block
- base_offset -> offset
- Remove mem_id concept
- Introduce graph::operand::Index instead of uint32_t
- Apply above changes to symbols of MemoryAllocator
- Apply above changes to tests of MemoryAllocator
- Append Logging
Signed-off-by: Yongseop Kim <yons.kim@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Tue, 30 Oct 2018 06:05:17 +0000 (15:05 +0900)]
Revise tizen build script (#3393)
- Revise tizen cross build script
- Revise tizen gbs build script
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
Removing ambigous/unused num_elem_processed_per_iteration variable (#3315)
This patch will remove unused num_elem_processed_per_iteration variable from CLGatherKernel.cpp
Signed-off-by: shubham <shub98.gupta@samsung.com>
Removing extra header files included in topkv2 files (#3372)
This patch will remove extra unnecessary headers in topkv2 op.
Also add CLScheduler.h which was getting included by CLArray
ClArray was including headers which were not required, hence it has been removed
Signed-off-by: shubham <shub98.gupta@samsung.com>
장지섭/동작제어Lab(SR)/Engineer/삼성전자 [Mon, 29 Oct 2018 11:30:47 +0000 (20:30 +0900)]
Add a generated test for HashTable_Lookup 4-D (#3381)
This commit add a generated test for HashTable_Lookup 4-D.
Signed-off-by: jiseob.jang <jiseob.jang@samsung.com>
김수진/동작제어Lab(SR)/Engineer/삼성전자 [Mon, 29 Oct 2018 11:29:43 +0000 (20:29 +0900)]
[neurun] Fix wrong ifdef guard (#3350)
This commit fixes ifdef guard in neurun.
Signed-off-by: sjsujinkim sjsujin.kim@samsung.com
Update nnapi delegate for LRN (#3270)
This patch will add nnapi delegation changes for LocalResponseNormalization.
Signed-off-by: shubham <shub98.gupta@samsung.com>
장지섭/동작제어Lab(SR)/Engineer/삼성전자 [Mon, 29 Oct 2018 06:33:15 +0000 (15:33 +0900)]
Add GeneratedTests of ReduceMax for 4-D (#3366)
This commit adds GeneratedTests of ReduceMax(TensorflowMax) for 4-D.
Signed-off-by: jiseob.jang <jiseob.jang@samsung.com>
Remove unnecessary headers from src/runtime/CL/functions/* (#3374)
This patch will remove unnecessary headers from src/runtime/CL/functions
Signed-off-by: shubham <shub98.gupta@samsung.com>
Remove unnecessary headers from arm_compute/runtime/CL/* (#3373)
This patch will remove unnecessary headers from arm_compute/runtime/CL files.
Signed-off-by: shubham <shub98.gupta@samsung.com>
Remove unnecessary headers from arm_compute/core/CL/* (#3371)
This patch will remove unnecessary headers included in the following files:
arm_compute/core/CL/kernels/CLActivationLayerExKernel.h
arm_compute/core/CL/kernels/CLArithmeticSubtractionExKernel.h
arm_compute/core/CL/kernels/CLEqualKernel.h
arm_compute/core/CL/kernels/CLGatherKernel.h
arm_compute/core/CL/kernels/CLNegKernel.h
arm_compute/core/CL/kernels/CLPixelWiseDivisionKernel.h
arm_compute/core/CL/kernels/CLReduceMaxKernel.h
arm_compute/core/CL/kernels/CLReductionMeanKernel.h
arm_compute/core/CL/kernels/CLSquaredDifferenceKernel.h
arm_compute/core/CL/kernels/CLStridedSliceKernel.h
Signed-off-by: shubham <shub98.gupta@samsung.com>
Updating addStage function declaration of Iplanbuilder (#3355)
This patch updates the declaration of addStage function of Iplanbuilder and it's associated comments.
Due to ambiguous 's' in doxygen comment, this change is made.
Signed-off-by: shubham <shub98.gupta@samsung.com>
Allign comment to have same formatting (#3369)
This patch will remove tab spaces and align the comment to have same formating style in same_padding function
Signed-off-by: shubham <shub98.gupta@samsung.com>
Support Padding accross all dimensions (#3305)
This patch will provide support to pad accross all dimensions of input tensor
Signed-off-by: shubham <shub98.gupta@samsung.com>
Correcting and adding arm_compute assertions in CL core/kernel files (#3312)
This patch will add more asertions required to be tested and corrects the wrong assertions that have been added.
Signed-off-by: shubham <shub98.gupta@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Fri, 26 Oct 2018 00:39:36 +0000 (09:39 +0900)]
Revise docker_build_cross_arm_ubuntu.sh (#3348)
Revise docker_build_cross_arm_ubuntu.sh
- Remove using nnfw_docker file
- Use docker-run and docker-shell command
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
Introduce ArgMax operation. (#3288)
This patch will introduce the ArgMax operation in NNFW.
Signed-off-by: Praveen D N <praveen.dn@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 25 Oct 2018 10:46:32 +0000 (19:46 +0900)]
[neurun] Fix Shape4D in LowerInfo order (#3339)
Fix Shape4D order bug (NHWC)
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
장지섭/동작제어Lab(SR)/Engineer/삼성전자 [Thu, 25 Oct 2018 10:42:59 +0000 (19:42 +0900)]
Fix padding bug of EmbeddingLookup (#3184)
This commit fixes padding bug of EmbeddingLookup.
Signed-off-by: jiseob.jang <jiseob.jang@samsung.com>
장지섭/동작제어Lab(SR)/Engineer/삼성전자 [Thu, 25 Oct 2018 09:43:05 +0000 (18:43 +0900)]
Support HashtableLookup greater than 2 dimensions (#3210)
This commit supports HashtableLookup greater than 2 dimensions.
Signed-off-by: jiseob.jang <jiseob.jang@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 25 Oct 2018 08:22:16 +0000 (17:22 +0900)]
Fix docker-shell command bug (#3336)
Need docker option -it for interactive
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
Remove unused headers and functions (#3316)
This patch will remove all unused headers and functions from following files:
CLActivationLayerExKernel.cpp
CLBatchToSpaceNDKernel.cpp
CLDepthToSpaceKernel.cpp
CLEqualKernel.cpp
CLGatherKernel.cpp
CLNegKernel.cpp
CLPixelWiseDivisionKernel.cpp
CLReduceMaxKernel.cpp
CLReduceSumKernel.cpp
CLReductionMeanKernel.cpp
CLSpaceToDepthKernel.cpp
CLSquaredDifferenceKernel.cpp
CLStridedSliceKernel.cpp
Signed-off-by: shubham <shub98.gupta@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 25 Oct 2018 07:18:43 +0000 (16:18 +0900)]
Check invalid format-checker exit (#3327)
Check invalid format-checker exit and return fail
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 25 Oct 2018 04:27:07 +0000 (13:27 +0900)]
Update docker-shell and configuration (#3324)
* Update docker-shell and configuration
Update docker-shell
- Use docker.configuration
Update docker.configuration
- Change default docker image name
- Fix tab to space
- Fix setting DOCKER_VOLUMES variable
- Enable using environment file
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
* Update environment variable setting
Remove unused environment setting and add download server setting
Setting missing quantization info (#3330)
This patch will pass scale and offset so as to set quantization info.
If required, this quantization info can be used by the operations
Signed-off-by: shubham <shub98.gupta@samsung.com>
Remove unused shape variables from L2Pool (#3329)
This patch will remove the unused shape variable from L2Pool.
Signed-off-by: shubham <shub98.gupta@samsung.com>
Add Generated Test for Testing broadcast DIV (#3296)
This patch adds a generated test for testing broadcast DIV.
This consist of testing broadcast div with tensors of shape 4D & 2D.
Added GTest : div_broadcast_float_4D_2D_nnfw
Signed-off-by: prasannar <prasanna.r@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Thu, 25 Oct 2018 00:09:52 +0000 (09:09 +0900)]
Fix python syntax error (#3326)
Fix python syntax error in select_operator.py
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
Change misleading FATAL error message from nnapi_delegate (#3307)
This patch changes misleading FATAL error message from nnapi_delegate for ```BatchToSpaceND```
Signed-off-by: prasannar <prasanna.r@samsung.com>
오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 [Wed, 24 Oct 2018 11:09:21 +0000 (20:09 +0900)]
Remove unused build script (#3323)
Remove unused two build script
- docker_build_cross_arm_neurun.sh (can build neurun using env file)
- docker_build_cross_arm_ubuntu_without_aclbuild.sh
Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
Remove passing pad gtest from skiplist (#3320)
This patch will remove pad gtest from skiplist
Signed-off-by: shubham <shub98.gupta@samsung.com>
이상규/동작제어Lab(SR)/Principal Engineer/삼성전자 [Wed, 24 Oct 2018 11:03:30 +0000 (20:03 +0900)]
Update `run_benchmark_op.sh` to enable neon properly (#3318)
Related Issue: #3262
`run_benchmark_op.sh` was written in the era of legacy runtime.
Currently we uses pure acl runtime as default.
Thus, I updated the way of enabling neon, and the labels.
Signed-off-by: Sanggyu Lee <sg5.lee@samsung.com>
Update assertions for BatchToSpaceND op in compile time. (#3286)
This patch updates the assertions required for BatchToSpaceND at compile time in PACL.
Signed-off-by: prasannar <prasanna.r@samsung.com>
Removing unused cstdint header from the runtime/CL/functions/CL* (#3311)
This patch will remove cstdint header from runtime/CL/functions/CL* since it is not getting used.
Signed-off-by: shubham <shub98.gupta@samsung.com>
Adding function call to validate_arg in CLGatherKernel.cpp (#3314)
This patch will add function call to validate_arg which was defined but was not getting used.
Signed-off-by: shubham <shub98.gupta@samsung.com>
Utility function to get ACL_PermutationVector from Runtime_PV (#3128)
For issue #3022
This commit adds a function that converts the Permutation vector of the runtime into the Permutation Vector of ACL
Signed-off-by: shubham <shub98.gupta@samsung.com>
Add CL Kernel Calls for SpaceToDepth from runtime. (#3267)
This patch adds CL Kernel Calls for SpaceToDepth from runtime
Signed-off-by: shubham <shub98.gupta@samsung.com>