| 
									
										
											  
											
												- build:
	- unify schema building in core and converter;
	- add more build script for android;
	- add linux build script for python;
- ops impl:
	- add floor mod support in binary;
	- use eltwise impl in add/max/sub/mul binary for optimization;
	- remove fake double support in cast;
	- fix 5d support for concat;
	- add adjX and adjY support for batch matmul;
	- optimize conv2d back prop filter;
	- add pad mode support for conv3d;
	- fix bug in conv2d & conv depthwise with very small feature map;
	- optimize binary without broacast;
	- add data types support for gather;
	- add gather ND support;
	- use uint8 data type in gather v2;
	- add transpose support for matmul;
	- add matrix band part;
	- add dim != 4 support for padding, reshape & tensor convert;
	- add pad type support for pool3d;
	- make ops based on TensorFlow Lite quantization optional;
	- add all & any support for reduction;
	- use type in parameter as output type in reduction;
	- add int support for unary;
	- add variable weight support for conv2d;
	- fix conv2d depthwise weights initialization;
	- fix type support for transpose;
	- fix grad outputs count for  reduce grad and reshape grad;
	- fix priorbox & detection output;
	- fix metal softmax error;
- python:
	- add runSessionWithCallBackInfo interface;
	- add max nodes limit (1400) for visualization tool;
	- fix save error in python3;
	- align default dim;
- convert:
	- add extra design for optimization;
	- add more post converting optimizers;
	- add caffe v1 weights blob support;
	- add cast, unary, conv transpose support for onnx model;
	- optimize batchnorm, conv with variable weights, prelu, reshape, slice, upsample for onnx model;
	- add cos/sin/atan/tan support for unary for tensorflow model;
	- add any/all support for reduction for tensorflow model;
	- add elu, conv3d, pool3d support for tensorflow model;
	- optimize argmax, batchnorm, concat, batch to space, conv with variable weights, prelu, slice for tensorflow model;
- others:
	- fix size computer lock;
	- fix thread pool deadlock;
	- add express & parameters in express;
	- rewrite blitter chooser without static map;
	- add tests for expr;
											
										 
											2019-10-29 13:37:26 +08:00
										 |  |  | //
 | 
					
						
							|  |  |  | //  MatrixBandTest.cpp
 | 
					
						
							|  |  |  | //  MNNTests
 | 
					
						
							|  |  |  | //
 | 
					
						
							|  |  |  | //  Created by MNN on 2019/09/17.
 | 
					
						
							|  |  |  | //  Copyright © 2018, Alibaba Group Holding Limited
 | 
					
						
							|  |  |  | //
 | 
					
						
							|  |  |  | 
 | 
					
						
							|  |  |  | /*
 | 
					
						
							|  |  |  |  Test Case From https://www.tensorflow.org/api_docs/cc/class/tensorflow/ops/matrix-band-part
 | 
					
						
							|  |  |  |  */ | 
					
						
							| 
									
										
										
										
											2019-12-27 22:16:57 +08:00
										 |  |  | #include <MNN/expr/ExprCreator.hpp>
 | 
					
						
							| 
									
										
											  
											
												- build:
	- unify schema building in core and converter;
	- add more build script for android;
	- add linux build script for python;
- ops impl:
	- add floor mod support in binary;
	- use eltwise impl in add/max/sub/mul binary for optimization;
	- remove fake double support in cast;
	- fix 5d support for concat;
	- add adjX and adjY support for batch matmul;
	- optimize conv2d back prop filter;
	- add pad mode support for conv3d;
	- fix bug in conv2d & conv depthwise with very small feature map;
	- optimize binary without broacast;
	- add data types support for gather;
	- add gather ND support;
	- use uint8 data type in gather v2;
	- add transpose support for matmul;
	- add matrix band part;
	- add dim != 4 support for padding, reshape & tensor convert;
	- add pad type support for pool3d;
	- make ops based on TensorFlow Lite quantization optional;
	- add all & any support for reduction;
	- use type in parameter as output type in reduction;
	- add int support for unary;
	- add variable weight support for conv2d;
	- fix conv2d depthwise weights initialization;
	- fix type support for transpose;
	- fix grad outputs count for  reduce grad and reshape grad;
	- fix priorbox & detection output;
	- fix metal softmax error;
- python:
	- add runSessionWithCallBackInfo interface;
	- add max nodes limit (1400) for visualization tool;
	- fix save error in python3;
	- align default dim;
- convert:
	- add extra design for optimization;
	- add more post converting optimizers;
	- add caffe v1 weights blob support;
	- add cast, unary, conv transpose support for onnx model;
	- optimize batchnorm, conv with variable weights, prelu, reshape, slice, upsample for onnx model;
	- add cos/sin/atan/tan support for unary for tensorflow model;
	- add any/all support for reduction for tensorflow model;
	- add elu, conv3d, pool3d support for tensorflow model;
	- optimize argmax, batchnorm, concat, batch to space, conv with variable weights, prelu, slice for tensorflow model;
- others:
	- fix size computer lock;
	- fix thread pool deadlock;
	- add express & parameters in express;
	- rewrite blitter chooser without static map;
	- add tests for expr;
											
										 
											2019-10-29 13:37:26 +08:00
										 |  |  | #include "MNNTestSuite.h"
 | 
					
						
							|  |  |  | #include "MNN_generated.h"
 | 
					
						
							|  |  |  | using namespace MNN::Express; | 
					
						
							|  |  |  | 
 | 
					
						
							|  |  |  | class MatrixBandTest : public MNNTestCase { | 
					
						
							|  |  |  | public: | 
					
						
							| 
									
										
										
										
											2021-06-11 17:17:13 +08:00
										 |  |  |     virtual bool run(int precision) { | 
					
						
							| 
									
										
											  
											
												- build:
	- unify schema building in core and converter;
	- add more build script for android;
	- add linux build script for python;
- ops impl:
	- add floor mod support in binary;
	- use eltwise impl in add/max/sub/mul binary for optimization;
	- remove fake double support in cast;
	- fix 5d support for concat;
	- add adjX and adjY support for batch matmul;
	- optimize conv2d back prop filter;
	- add pad mode support for conv3d;
	- fix bug in conv2d & conv depthwise with very small feature map;
	- optimize binary without broacast;
	- add data types support for gather;
	- add gather ND support;
	- use uint8 data type in gather v2;
	- add transpose support for matmul;
	- add matrix band part;
	- add dim != 4 support for padding, reshape & tensor convert;
	- add pad type support for pool3d;
	- make ops based on TensorFlow Lite quantization optional;
	- add all & any support for reduction;
	- use type in parameter as output type in reduction;
	- add int support for unary;
	- add variable weight support for conv2d;
	- fix conv2d depthwise weights initialization;
	- fix type support for transpose;
	- fix grad outputs count for  reduce grad and reshape grad;
	- fix priorbox & detection output;
	- fix metal softmax error;
- python:
	- add runSessionWithCallBackInfo interface;
	- add max nodes limit (1400) for visualization tool;
	- fix save error in python3;
	- align default dim;
- convert:
	- add extra design for optimization;
	- add more post converting optimizers;
	- add caffe v1 weights blob support;
	- add cast, unary, conv transpose support for onnx model;
	- optimize batchnorm, conv with variable weights, prelu, reshape, slice, upsample for onnx model;
	- add cos/sin/atan/tan support for unary for tensorflow model;
	- add any/all support for reduction for tensorflow model;
	- add elu, conv3d, pool3d support for tensorflow model;
	- optimize argmax, batchnorm, concat, batch to space, conv with variable weights, prelu, slice for tensorflow model;
- others:
	- fix size computer lock;
	- fix thread pool deadlock;
	- add express & parameters in express;
	- rewrite blitter chooser without static map;
	- add tests for expr;
											
										 
											2019-10-29 13:37:26 +08:00
										 |  |  |         std::unique_ptr<MNN::OpT> MatrixBandOp(new MNN::OpT); | 
					
						
							| 
									
										
										
										
											2020-11-05 16:41:56 +08:00
										 |  |  |         MatrixBandOp->type        = MNN::OpType_MatrixBandPart; | 
					
						
							|  |  |  |         auto matrix               = _Input({4, 4}, NHWC, halide_type_of<float>()); | 
					
						
							|  |  |  |         auto lower                = _Input({}, NHWC, halide_type_of<int32_t>()); | 
					
						
							|  |  |  |         auto upper                = _Input({}, NHWC, halide_type_of<int32_t>()); | 
					
						
							|  |  |  |         auto y                    = Variable::create(Expr::create(MatrixBandOp.get(), {matrix, lower, upper})); | 
					
						
							|  |  |  |         std::vector<float> values = {0.0f,  1.0f,  2.0f, 3.0f, -1.0f, 0.0f,  1.0f,  2.0f, | 
					
						
							|  |  |  |                                      -2.0f, -1.0f, 0.0f, 1.0f, -3.0f, -2.0f, -1.0f, 0.0f}; | 
					
						
							| 
									
										
											  
											
												- build:
	- unify schema building in core and converter;
	- add more build script for android;
	- add linux build script for python;
- ops impl:
	- add floor mod support in binary;
	- use eltwise impl in add/max/sub/mul binary for optimization;
	- remove fake double support in cast;
	- fix 5d support for concat;
	- add adjX and adjY support for batch matmul;
	- optimize conv2d back prop filter;
	- add pad mode support for conv3d;
	- fix bug in conv2d & conv depthwise with very small feature map;
	- optimize binary without broacast;
	- add data types support for gather;
	- add gather ND support;
	- use uint8 data type in gather v2;
	- add transpose support for matmul;
	- add matrix band part;
	- add dim != 4 support for padding, reshape & tensor convert;
	- add pad type support for pool3d;
	- make ops based on TensorFlow Lite quantization optional;
	- add all & any support for reduction;
	- use type in parameter as output type in reduction;
	- add int support for unary;
	- add variable weight support for conv2d;
	- fix conv2d depthwise weights initialization;
	- fix type support for transpose;
	- fix grad outputs count for  reduce grad and reshape grad;
	- fix priorbox & detection output;
	- fix metal softmax error;
- python:
	- add runSessionWithCallBackInfo interface;
	- add max nodes limit (1400) for visualization tool;
	- fix save error in python3;
	- align default dim;
- convert:
	- add extra design for optimization;
	- add more post converting optimizers;
	- add caffe v1 weights blob support;
	- add cast, unary, conv transpose support for onnx model;
	- optimize batchnorm, conv with variable weights, prelu, reshape, slice, upsample for onnx model;
	- add cos/sin/atan/tan support for unary for tensorflow model;
	- add any/all support for reduction for tensorflow model;
	- add elu, conv3d, pool3d support for tensorflow model;
	- optimize argmax, batchnorm, concat, batch to space, conv with variable weights, prelu, slice for tensorflow model;
- others:
	- fix size computer lock;
	- fix thread pool deadlock;
	- add express & parameters in express;
	- rewrite blitter chooser without static map;
	- add tests for expr;
											
										 
											2019-10-29 13:37:26 +08:00
										 |  |  |         ::memcpy(matrix->writeMap<float>(), values.data(), values.size() * sizeof(float)); | 
					
						
							|  |  |  |         { | 
					
						
							|  |  |  |             lower->writeMap<int>()[0] = 1; | 
					
						
							|  |  |  |             upper->writeMap<int>()[0] = -1; | 
					
						
							|  |  |  |             { | 
					
						
							|  |  |  |                 auto yPtr = y->readMap<float>(); | 
					
						
							| 
									
										
										
										
											2020-11-05 16:41:56 +08:00
										 |  |  |                 for (int h = 0; h < 4; ++h) { | 
					
						
							|  |  |  |                     for (int w = 0; w < 4; ++w) { | 
					
						
							|  |  |  |                         auto computed = yPtr[4 * h + w]; | 
					
						
							| 
									
										
											  
											
												- build:
	- unify schema building in core and converter;
	- add more build script for android;
	- add linux build script for python;
- ops impl:
	- add floor mod support in binary;
	- use eltwise impl in add/max/sub/mul binary for optimization;
	- remove fake double support in cast;
	- fix 5d support for concat;
	- add adjX and adjY support for batch matmul;
	- optimize conv2d back prop filter;
	- add pad mode support for conv3d;
	- fix bug in conv2d & conv depthwise with very small feature map;
	- optimize binary without broacast;
	- add data types support for gather;
	- add gather ND support;
	- use uint8 data type in gather v2;
	- add transpose support for matmul;
	- add matrix band part;
	- add dim != 4 support for padding, reshape & tensor convert;
	- add pad type support for pool3d;
	- make ops based on TensorFlow Lite quantization optional;
	- add all & any support for reduction;
	- use type in parameter as output type in reduction;
	- add int support for unary;
	- add variable weight support for conv2d;
	- fix conv2d depthwise weights initialization;
	- fix type support for transpose;
	- fix grad outputs count for  reduce grad and reshape grad;
	- fix priorbox & detection output;
	- fix metal softmax error;
- python:
	- add runSessionWithCallBackInfo interface;
	- add max nodes limit (1400) for visualization tool;
	- fix save error in python3;
	- align default dim;
- convert:
	- add extra design for optimization;
	- add more post converting optimizers;
	- add caffe v1 weights blob support;
	- add cast, unary, conv transpose support for onnx model;
	- optimize batchnorm, conv with variable weights, prelu, reshape, slice, upsample for onnx model;
	- add cos/sin/atan/tan support for unary for tensorflow model;
	- add any/all support for reduction for tensorflow model;
	- add elu, conv3d, pool3d support for tensorflow model;
	- optimize argmax, batchnorm, concat, batch to space, conv with variable weights, prelu, slice for tensorflow model;
- others:
	- fix size computer lock;
	- fix thread pool deadlock;
	- add express & parameters in express;
	- rewrite blitter chooser without static map;
	- add tests for expr;
											
										 
											2019-10-29 13:37:26 +08:00
										 |  |  |                         auto expected = 0.0f; | 
					
						
							| 
									
										
										
										
											2020-11-05 16:41:56 +08:00
										 |  |  |                         if (h - w <= 1) { | 
					
						
							|  |  |  |                             expected = values[4 * h + w]; | 
					
						
							| 
									
										
											  
											
												- build:
	- unify schema building in core and converter;
	- add more build script for android;
	- add linux build script for python;
- ops impl:
	- add floor mod support in binary;
	- use eltwise impl in add/max/sub/mul binary for optimization;
	- remove fake double support in cast;
	- fix 5d support for concat;
	- add adjX and adjY support for batch matmul;
	- optimize conv2d back prop filter;
	- add pad mode support for conv3d;
	- fix bug in conv2d & conv depthwise with very small feature map;
	- optimize binary without broacast;
	- add data types support for gather;
	- add gather ND support;
	- use uint8 data type in gather v2;
	- add transpose support for matmul;
	- add matrix band part;
	- add dim != 4 support for padding, reshape & tensor convert;
	- add pad type support for pool3d;
	- make ops based on TensorFlow Lite quantization optional;
	- add all & any support for reduction;
	- use type in parameter as output type in reduction;
	- add int support for unary;
	- add variable weight support for conv2d;
	- fix conv2d depthwise weights initialization;
	- fix type support for transpose;
	- fix grad outputs count for  reduce grad and reshape grad;
	- fix priorbox & detection output;
	- fix metal softmax error;
- python:
	- add runSessionWithCallBackInfo interface;
	- add max nodes limit (1400) for visualization tool;
	- fix save error in python3;
	- align default dim;
- convert:
	- add extra design for optimization;
	- add more post converting optimizers;
	- add caffe v1 weights blob support;
	- add cast, unary, conv transpose support for onnx model;
	- optimize batchnorm, conv with variable weights, prelu, reshape, slice, upsample for onnx model;
	- add cos/sin/atan/tan support for unary for tensorflow model;
	- add any/all support for reduction for tensorflow model;
	- add elu, conv3d, pool3d support for tensorflow model;
	- optimize argmax, batchnorm, concat, batch to space, conv with variable weights, prelu, slice for tensorflow model;
- others:
	- fix size computer lock;
	- fix thread pool deadlock;
	- add express & parameters in express;
	- rewrite blitter chooser without static map;
	- add tests for expr;
											
										 
											2019-10-29 13:37:26 +08:00
										 |  |  |                         } | 
					
						
							|  |  |  |                         if (computed != expected) { | 
					
						
							|  |  |  |                             FUNC_PRINT(1); | 
					
						
							|  |  |  |                             return false; | 
					
						
							|  |  |  |                         } | 
					
						
							|  |  |  |                     } | 
					
						
							|  |  |  |                 } | 
					
						
							|  |  |  |             } | 
					
						
							|  |  |  |         } | 
					
						
							|  |  |  |         { | 
					
						
							|  |  |  |             lower->writeMap<int>()[0] = 2; | 
					
						
							|  |  |  |             upper->writeMap<int>()[0] = 1; | 
					
						
							|  |  |  |             { | 
					
						
							|  |  |  |                 auto yPtr = y->readMap<float>(); | 
					
						
							| 
									
										
										
										
											2020-11-05 16:41:56 +08:00
										 |  |  |                 for (int h = 0; h < 4; ++h) { | 
					
						
							|  |  |  |                     for (int w = 0; w < 4; ++w) { | 
					
						
							|  |  |  |                         auto computed = yPtr[4 * h + w]; | 
					
						
							| 
									
										
											  
											
												- build:
	- unify schema building in core and converter;
	- add more build script for android;
	- add linux build script for python;
- ops impl:
	- add floor mod support in binary;
	- use eltwise impl in add/max/sub/mul binary for optimization;
	- remove fake double support in cast;
	- fix 5d support for concat;
	- add adjX and adjY support for batch matmul;
	- optimize conv2d back prop filter;
	- add pad mode support for conv3d;
	- fix bug in conv2d & conv depthwise with very small feature map;
	- optimize binary without broacast;
	- add data types support for gather;
	- add gather ND support;
	- use uint8 data type in gather v2;
	- add transpose support for matmul;
	- add matrix band part;
	- add dim != 4 support for padding, reshape & tensor convert;
	- add pad type support for pool3d;
	- make ops based on TensorFlow Lite quantization optional;
	- add all & any support for reduction;
	- use type in parameter as output type in reduction;
	- add int support for unary;
	- add variable weight support for conv2d;
	- fix conv2d depthwise weights initialization;
	- fix type support for transpose;
	- fix grad outputs count for  reduce grad and reshape grad;
	- fix priorbox & detection output;
	- fix metal softmax error;
- python:
	- add runSessionWithCallBackInfo interface;
	- add max nodes limit (1400) for visualization tool;
	- fix save error in python3;
	- align default dim;
- convert:
	- add extra design for optimization;
	- add more post converting optimizers;
	- add caffe v1 weights blob support;
	- add cast, unary, conv transpose support for onnx model;
	- optimize batchnorm, conv with variable weights, prelu, reshape, slice, upsample for onnx model;
	- add cos/sin/atan/tan support for unary for tensorflow model;
	- add any/all support for reduction for tensorflow model;
	- add elu, conv3d, pool3d support for tensorflow model;
	- optimize argmax, batchnorm, concat, batch to space, conv with variable weights, prelu, slice for tensorflow model;
- others:
	- fix size computer lock;
	- fix thread pool deadlock;
	- add express & parameters in express;
	- rewrite blitter chooser without static map;
	- add tests for expr;
											
										 
											2019-10-29 13:37:26 +08:00
										 |  |  |                         auto expected = 0.0f; | 
					
						
							| 
									
										
										
										
											2020-11-05 16:41:56 +08:00
										 |  |  |                         if ((h - w) <= 2 && (w - h) <= 1) { | 
					
						
							|  |  |  |                             expected = values[4 * h + w]; | 
					
						
							| 
									
										
											  
											
												- build:
	- unify schema building in core and converter;
	- add more build script for android;
	- add linux build script for python;
- ops impl:
	- add floor mod support in binary;
	- use eltwise impl in add/max/sub/mul binary for optimization;
	- remove fake double support in cast;
	- fix 5d support for concat;
	- add adjX and adjY support for batch matmul;
	- optimize conv2d back prop filter;
	- add pad mode support for conv3d;
	- fix bug in conv2d & conv depthwise with very small feature map;
	- optimize binary without broacast;
	- add data types support for gather;
	- add gather ND support;
	- use uint8 data type in gather v2;
	- add transpose support for matmul;
	- add matrix band part;
	- add dim != 4 support for padding, reshape & tensor convert;
	- add pad type support for pool3d;
	- make ops based on TensorFlow Lite quantization optional;
	- add all & any support for reduction;
	- use type in parameter as output type in reduction;
	- add int support for unary;
	- add variable weight support for conv2d;
	- fix conv2d depthwise weights initialization;
	- fix type support for transpose;
	- fix grad outputs count for  reduce grad and reshape grad;
	- fix priorbox & detection output;
	- fix metal softmax error;
- python:
	- add runSessionWithCallBackInfo interface;
	- add max nodes limit (1400) for visualization tool;
	- fix save error in python3;
	- align default dim;
- convert:
	- add extra design for optimization;
	- add more post converting optimizers;
	- add caffe v1 weights blob support;
	- add cast, unary, conv transpose support for onnx model;
	- optimize batchnorm, conv with variable weights, prelu, reshape, slice, upsample for onnx model;
	- add cos/sin/atan/tan support for unary for tensorflow model;
	- add any/all support for reduction for tensorflow model;
	- add elu, conv3d, pool3d support for tensorflow model;
	- optimize argmax, batchnorm, concat, batch to space, conv with variable weights, prelu, slice for tensorflow model;
- others:
	- fix size computer lock;
	- fix thread pool deadlock;
	- add express & parameters in express;
	- rewrite blitter chooser without static map;
	- add tests for expr;
											
										 
											2019-10-29 13:37:26 +08:00
										 |  |  |                         } | 
					
						
							|  |  |  |                         if (computed != expected) { | 
					
						
							|  |  |  |                             FUNC_PRINT(1); | 
					
						
							|  |  |  |                             return false; | 
					
						
							|  |  |  |                         } | 
					
						
							|  |  |  |                     } | 
					
						
							|  |  |  |                 } | 
					
						
							|  |  |  |             } | 
					
						
							|  |  |  |         } | 
					
						
							|  |  |  |         { | 
					
						
							|  |  |  |             matrix->resize({3, 5, 5}); | 
					
						
							|  |  |  |             auto matrixPtr = matrix->writeMap<float>(); | 
					
						
							| 
									
										
										
										
											2020-11-05 16:41:56 +08:00
										 |  |  |             for (int i = 0; i < matrix->getInfo()->size; ++i) { | 
					
						
							| 
									
										
											  
											
												- build:
	- unify schema building in core and converter;
	- add more build script for android;
	- add linux build script for python;
- ops impl:
	- add floor mod support in binary;
	- use eltwise impl in add/max/sub/mul binary for optimization;
	- remove fake double support in cast;
	- fix 5d support for concat;
	- add adjX and adjY support for batch matmul;
	- optimize conv2d back prop filter;
	- add pad mode support for conv3d;
	- fix bug in conv2d & conv depthwise with very small feature map;
	- optimize binary without broacast;
	- add data types support for gather;
	- add gather ND support;
	- use uint8 data type in gather v2;
	- add transpose support for matmul;
	- add matrix band part;
	- add dim != 4 support for padding, reshape & tensor convert;
	- add pad type support for pool3d;
	- make ops based on TensorFlow Lite quantization optional;
	- add all & any support for reduction;
	- use type in parameter as output type in reduction;
	- add int support for unary;
	- add variable weight support for conv2d;
	- fix conv2d depthwise weights initialization;
	- fix type support for transpose;
	- fix grad outputs count for  reduce grad and reshape grad;
	- fix priorbox & detection output;
	- fix metal softmax error;
- python:
	- add runSessionWithCallBackInfo interface;
	- add max nodes limit (1400) for visualization tool;
	- fix save error in python3;
	- align default dim;
- convert:
	- add extra design for optimization;
	- add more post converting optimizers;
	- add caffe v1 weights blob support;
	- add cast, unary, conv transpose support for onnx model;
	- optimize batchnorm, conv with variable weights, prelu, reshape, slice, upsample for onnx model;
	- add cos/sin/atan/tan support for unary for tensorflow model;
	- add any/all support for reduction for tensorflow model;
	- add elu, conv3d, pool3d support for tensorflow model;
	- optimize argmax, batchnorm, concat, batch to space, conv with variable weights, prelu, slice for tensorflow model;
- others:
	- fix size computer lock;
	- fix thread pool deadlock;
	- add express & parameters in express;
	- rewrite blitter chooser without static map;
	- add tests for expr;
											
										 
											2019-10-29 13:37:26 +08:00
										 |  |  |                 matrixPtr[i] = (float)i; | 
					
						
							|  |  |  |             } | 
					
						
							|  |  |  |             lower->writeMap<int>()[0] = 2; | 
					
						
							|  |  |  |             upper->writeMap<int>()[0] = 1; | 
					
						
							| 
									
										
										
										
											2020-11-05 16:41:56 +08:00
										 |  |  |             auto yPtr                 = y->readMap<float>(); | 
					
						
							|  |  |  |             for (int z = 0; z < 3; ++z) { | 
					
						
							|  |  |  |                 for (int h = 0; h < 5; ++h) { | 
					
						
							|  |  |  |                     for (int w = 0; w < 5; ++w) { | 
					
						
							|  |  |  |                         auto index    = w + 5 * h + 5 * 5 * z; | 
					
						
							| 
									
										
											  
											
												- build:
	- unify schema building in core and converter;
	- add more build script for android;
	- add linux build script for python;
- ops impl:
	- add floor mod support in binary;
	- use eltwise impl in add/max/sub/mul binary for optimization;
	- remove fake double support in cast;
	- fix 5d support for concat;
	- add adjX and adjY support for batch matmul;
	- optimize conv2d back prop filter;
	- add pad mode support for conv3d;
	- fix bug in conv2d & conv depthwise with very small feature map;
	- optimize binary without broacast;
	- add data types support for gather;
	- add gather ND support;
	- use uint8 data type in gather v2;
	- add transpose support for matmul;
	- add matrix band part;
	- add dim != 4 support for padding, reshape & tensor convert;
	- add pad type support for pool3d;
	- make ops based on TensorFlow Lite quantization optional;
	- add all & any support for reduction;
	- use type in parameter as output type in reduction;
	- add int support for unary;
	- add variable weight support for conv2d;
	- fix conv2d depthwise weights initialization;
	- fix type support for transpose;
	- fix grad outputs count for  reduce grad and reshape grad;
	- fix priorbox & detection output;
	- fix metal softmax error;
- python:
	- add runSessionWithCallBackInfo interface;
	- add max nodes limit (1400) for visualization tool;
	- fix save error in python3;
	- align default dim;
- convert:
	- add extra design for optimization;
	- add more post converting optimizers;
	- add caffe v1 weights blob support;
	- add cast, unary, conv transpose support for onnx model;
	- optimize batchnorm, conv with variable weights, prelu, reshape, slice, upsample for onnx model;
	- add cos/sin/atan/tan support for unary for tensorflow model;
	- add any/all support for reduction for tensorflow model;
	- add elu, conv3d, pool3d support for tensorflow model;
	- optimize argmax, batchnorm, concat, batch to space, conv with variable weights, prelu, slice for tensorflow model;
- others:
	- fix size computer lock;
	- fix thread pool deadlock;
	- add express & parameters in express;
	- rewrite blitter chooser without static map;
	- add tests for expr;
											
										 
											2019-10-29 13:37:26 +08:00
										 |  |  |                         auto computed = yPtr[index]; | 
					
						
							|  |  |  |                         auto expected = 0.0f; | 
					
						
							| 
									
										
										
										
											2020-11-05 16:41:56 +08:00
										 |  |  |                         if ((h - w) <= 2 && (w - h) <= 1) { | 
					
						
							| 
									
										
											  
											
												- build:
	- unify schema building in core and converter;
	- add more build script for android;
	- add linux build script for python;
- ops impl:
	- add floor mod support in binary;
	- use eltwise impl in add/max/sub/mul binary for optimization;
	- remove fake double support in cast;
	- fix 5d support for concat;
	- add adjX and adjY support for batch matmul;
	- optimize conv2d back prop filter;
	- add pad mode support for conv3d;
	- fix bug in conv2d & conv depthwise with very small feature map;
	- optimize binary without broacast;
	- add data types support for gather;
	- add gather ND support;
	- use uint8 data type in gather v2;
	- add transpose support for matmul;
	- add matrix band part;
	- add dim != 4 support for padding, reshape & tensor convert;
	- add pad type support for pool3d;
	- make ops based on TensorFlow Lite quantization optional;
	- add all & any support for reduction;
	- use type in parameter as output type in reduction;
	- add int support for unary;
	- add variable weight support for conv2d;
	- fix conv2d depthwise weights initialization;
	- fix type support for transpose;
	- fix grad outputs count for  reduce grad and reshape grad;
	- fix priorbox & detection output;
	- fix metal softmax error;
- python:
	- add runSessionWithCallBackInfo interface;
	- add max nodes limit (1400) for visualization tool;
	- fix save error in python3;
	- align default dim;
- convert:
	- add extra design for optimization;
	- add more post converting optimizers;
	- add caffe v1 weights blob support;
	- add cast, unary, conv transpose support for onnx model;
	- optimize batchnorm, conv with variable weights, prelu, reshape, slice, upsample for onnx model;
	- add cos/sin/atan/tan support for unary for tensorflow model;
	- add any/all support for reduction for tensorflow model;
	- add elu, conv3d, pool3d support for tensorflow model;
	- optimize argmax, batchnorm, concat, batch to space, conv with variable weights, prelu, slice for tensorflow model;
- others:
	- fix size computer lock;
	- fix thread pool deadlock;
	- add express & parameters in express;
	- rewrite blitter chooser without static map;
	- add tests for expr;
											
										 
											2019-10-29 13:37:26 +08:00
										 |  |  |                             expected = (float)(index); | 
					
						
							|  |  |  |                         } | 
					
						
							|  |  |  |                         if (computed != expected) { | 
					
						
							|  |  |  |                             FUNC_PRINT(1); | 
					
						
							|  |  |  |                             return false; | 
					
						
							|  |  |  |                         } | 
					
						
							|  |  |  |                     } | 
					
						
							|  |  |  |                 } | 
					
						
							|  |  |  |             } | 
					
						
							|  |  |  |         } | 
					
						
							|  |  |  |         return true; | 
					
						
							|  |  |  |     } | 
					
						
							|  |  |  | }; | 
					
						
							|  |  |  | MNNTestSuiteRegister(MatrixBandTest, "expr/MatrixBand"); |