24.02.1
|
Go to the documentation of this file.
44 using namespace misc::shape_calculator;
52 const ITensorInfo *
dst,
53 const Size2D &convolved_dims,
62 if (
dst->total_size() != 0)
69 "Col2Im output's data layout must always be NCHW");
75 std::pair<Status, Window>
84 constexpr
unsigned int num_elems_read_per_iteration = 8;
90 AccessWindowHorizontal input_access(
src, 0, num_elems_read_per_iteration);
95 return std::make_pair(err, win);
107 const Size2D &convolved_dims,
132 IClKernel::configure_internal(win_config.second);
135 _config_id =
"col2im_";
151 const Size2D &convolved_dims,
166 bool is_collapsed =
false;
167 bool is_collapsed_out =
false;
185 unsigned int idx = 0;
@ NCHW
Num samples, channels, height, width.
std::string to_string(T &&value)
Convert integer and float values to string.
SimpleTensor< float > src
const StringSet & options() const
Gets the current options list set.
Window calculate_max_window(const ValidRegion &valid_region, const Steps &steps, bool skip_border, BorderSize border_size)
void configure(const CLCompileContext &compile_context, ITensorInfo *src, ITensorInfo *dst, const Size2D &convolved_dims, unsigned int num_groups=1)
Set the input and output of the kernel.
@ QASYMM8
quantized, asymmetric fixed-point 8-bit number unsigned
std::string lower_string(const std::string &val)
Lower a given string.
Status validate_arguments(const ITensorInfo *src, const ITensorInfo *weights, const ITensorInfo *dst, const PadStrideInfo &conv_info)
#define ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL(k)
Window collapse_if_possible(const Window &full_window, size_t first, size_t last, bool *has_collapsed=nullptr) const
Collapse the dimensions between first and last if possible.
Class for specifying the size of an image or rectangle.
void use_tensor_dimensions(const TensorShape &shape, size_t first_dimension=Window::DimX)
Use the tensor's dimensions to fill the window dimensions.
bool update_window_and_padding(Window &win, Ts &&...patterns)
Update window and padding size for each of the access patterns.
ITensor * get_tensor(int id)
Get tensor of a given id from the pac.
src_info set_data_layout(data_layout)
const std::string & string_from_data_type(DataType dt)
Convert a data type identity into a string.
#define ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_DATA_TYPES(...)
#define ARM_COMPUTE_RETURN_ERROR_ON_DATA_TYPE_CHANNEL_NOT_IN(t, c,...)
Manages all the OpenCL kernels compilation and caching, provides accessors for the OpenCL Context.
#define ARM_COMPUTE_RETURN_ON_ERROR(status)
Checks if a status contains an error and returns it.
@ RUNTIME_ERROR
Generic runtime error.
#define ARM_COMPUTE_ERROR_ON_NULLPTR(...)
#define ARM_COMPUTE_ERROR_ON(cond)
If the condition is true then an error message is printed and an exception thrown.
const ITensor * get_const_tensor(int id) const
Get constant tensor of a given id.
#define ARM_COMPUTE_ERROR_THROW_ON(status)
size_t width
Width of the image region or rectangle.
void add_option(std::string option)
Adds option to the existing build option list.
bool auto_init_if_empty(ITensorInfo &info, const TensorShape &shape, int num_channels, DataType data_type, QuantizationInfo quantization_info=QuantizationInfo())
Auto initialize the tensor info (shape, number of channels and data type) if the current assignment i...
#define ARM_COMPUTE_RETURN_ERROR_ON_F16_UNSUPPORTED(tensor)
cl::Kernel create_kernel(const CLCompileContext &ctx, const std::string &kernel_name, const std::set< std::string > &build_opts=std::set< std::string >())
Creates an opencl kernel using a compile context.
@ QASYMM8_SIGNED
quantized, asymmetric fixed-point 8-bit number signed
bool slide_window_slice_3D(Window &slice) const
Slide the passed 3D window slice.
std::pair< Status, Window > validate_and_configure_window(ITensorInfo *src, ITensorInfo *dst)
Wrapper to configure the Khronos OpenCL C++ header.
void run_op(ITensorPack &tensors, const Window &window, cl::CommandQueue &queue) override
Enqueue the OpenCL kernel to process the given window on the passed OpenCL command queue.
void add_4D_tensor_argument(unsigned int &idx, const ICLTensor *tensor, const Window &window)
Add the passed 4D tensor's parameters to the object's kernel's arguments starting from the index idx.
Window first_slice_window_3D() const
First 3D slice of the window.
const Window & window() const
The maximum window the kernel can be executed on.
#define ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_QUANTIZATION_INFO(...)
std::string get_cl_type_from_data_type(const DataType &dt)
Translates a tensor data type to the appropriate OpenCL type.
void add_3D_tensor_argument(unsigned int &idx, const ICLTensor *tensor, const Window &window)
Add the passed 3D tensor's parameters to the object's kernel's arguments starting from the index idx.
const unsigned int num_groups
Describe a multidimensional execution window.
@ ELEMENTWISE
Elementwise CL kernel type.
#define ARM_COMPUTE_RETURN_ERROR_ON_MSG(cond, msg)
If the condition is true, an error is returned.
Copyright (c) 2017-2024 Arm Limited.
#define ARM_COMPUTE_RETURN_ERROR_ON_MISMATCHING_DIMENSIONS(...)
ClCol2ImKernel()
Default constructor.
static Status validate(const ITensorInfo *src, const ITensorInfo *dst, const Size2D &convolved_dims, unsigned int num_groups=1)
Static function to check if given info will lead to a valid configuration.
@ F16
16-bit floating-point number
static constexpr size_t DimZ
Alias for dimension 2 also known as Z dimension.
cl::NDRange lws_hint() const
Return the Local-Workgroup-Size hint.
#define ARM_COMPUTE_RETURN_ERROR_ON_NULLPTR(...)
Store the tensor's metadata.
Window first_slice_window_4D() const
First 4D slice of the window.
@ F32
32-bit floating-point number
#define ARM_COMPUTE_CREATE_ERROR(error_code, msg)
Creates an error with a given message.
#define ARM_COMPUTE_ERROR_ON_MISMATCHING_WINDOWS(f, w)
bool slide_window_slice_4D(Window &slice) const
Slide the passed 4D window slice.
TensorShape compute_col2im_shape(const ITensorInfo &input, const Size2D &convolved_dims, bool batch_size_on_z, unsigned int num_groups=1)
Calculate the Col2Im shape.
DataType
Available data types.
SimpleTensor< T > slice(const SimpleTensor< T > &src, Coordinates starts, Coordinates ends)
void enqueue(cl::CommandQueue &queue, ICLKernel &kernel, const Window &window, const cl::NDRange &lws_hint=CLKernelLibrary::get().default_ndrange(), bool use_dummy_work_items=false)
Add the kernel to the command queue with the given window.