- 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
|
|
|
|
|
|
|
//
|
|
|
|
// CPUGatherND.cpp
|
|
|
|
// MNN
|
|
|
|
//
|
|
|
|
// Created by MNN on 2019/09/11.
|
|
|
|
// Copyright © 2018, Alibaba Group Holding Limited
|
|
|
|
//
|
|
|
|
|
|
|
|
/*Ref:
|
|
|
|
https://github.com/tensorflow/tensorflow/tree/master/tensorflow/lite/kernels/internal/reference/reference_ops.h
|
|
|
|
*/
|
|
|
|
|
2019-12-27 22:16:57 +08:00
|
|
|
#include "backend/cpu/CPUGatherND.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 <string.h>
|
|
|
|
|
|
|
|
namespace MNN {
|
|
|
|
ErrorCode CPUGatherND::onResize(const std::vector<Tensor *> &inputs, const std::vector<Tensor *> &outputs) {
|
|
|
|
auto params = inputs[0];
|
|
|
|
auto indice = inputs[1];
|
|
|
|
mSliceN = 1;
|
|
|
|
mSliceSize = 1;
|
|
|
|
for (int i=0; i<indice->dimensions()-1; ++i) {
|
|
|
|
mSliceN *= indice->length(i);
|
|
|
|
}
|
|
|
|
auto indiceNd = indice->length(indice->dimensions()-1);
|
|
|
|
mDimsToCount.resize(indiceNd);
|
|
|
|
for (int i=indiceNd; i<params->dimensions(); ++i) {
|
|
|
|
mSliceSize *= params->length(i);
|
|
|
|
}
|
|
|
|
auto paramSize = params->elementSize();
|
|
|
|
for (int i=0; i<indiceNd; ++i) {
|
|
|
|
mDimsToCount[i] = paramSize / params->length(i);
|
|
|
|
paramSize = mDimsToCount[i];
|
|
|
|
}
|
|
|
|
mDimsToCount.resize(indiceNd);
|
|
|
|
return NO_ERROR;
|
|
|
|
}
|
|
|
|
|
|
|
|
ErrorCode CPUGatherND::onExecute(const std::vector<Tensor *> &inputs, const std::vector<Tensor *> &outputs) {
|
|
|
|
auto params = inputs[0];
|
|
|
|
auto indice = inputs[1];
|
|
|
|
auto indiceNd = indice->length(indice->dimensions()-1);
|
|
|
|
auto indiceData = indice->host<int32_t>();
|
|
|
|
auto output = outputs[0];
|
|
|
|
auto bytes = output->getType().bytes();
|
|
|
|
|
|
|
|
for (int i=0; i<mSliceN; ++i) {
|
|
|
|
int fromPos = 0;
|
|
|
|
for (int j=0; j<indiceNd; ++j) {
|
|
|
|
fromPos += mDimsToCount[j] * indiceData[i*indiceNd + j];
|
|
|
|
}
|
|
|
|
::memcpy(output->host<uint8_t>() + bytes * i * mSliceSize, params->host<uint8_t>() + bytes * fromPos, bytes * mSliceSize);
|
|
|
|
}
|
|
|
|
return NO_ERROR;
|
|
|
|
}
|
|
|
|
|
|
|
|
class CPUGatherNDCreator : public CPUBackend::Creator {
|
|
|
|
public:
|
|
|
|
virtual Execution *onCreate(const std::vector<Tensor *> &inputs, const std::vector<Tensor *> &outputs,
|
|
|
|
const MNN::Op *op, Backend *backend) const override {
|
|
|
|
return new CPUGatherND(backend);
|
|
|
|
}
|
|
|
|
};
|
|
|
|
|
|
|
|
REGISTER_CPU_OP_CREATOR(CPUGatherNDCreator, OpType_GatherND);
|
|
|
|
|
|
|
|
} // namespace MNN
|