#include <forest_model.hpp>

Collaboration diagram for ML::experimental::fil::forest_model:

Public Member Functions
	forest_model (decision_forest_variant &&forest=decision_forest_variant{})

auto	num_features ()

auto	num_outputs ()

auto	num_trees ()

auto	has_vector_leaves ()

auto	row_postprocessing ()

void	set_row_postprocessing (row_op val)

auto	elem_postprocessing ()

auto	memory_type ()

auto	device_index ()

auto	is_double_precision ()

template<typename io_t >
void	predict (raft_proto::buffer< io_t > &output, raft_proto::buffer< io_t > const &input, raft_proto::cuda_stream stream=raft_proto::cuda_stream{}, infer_kind predict_type=infer_kind::default_kind, std::optional< index_type > specified_chunk_size=std::nullopt)

template<typename io_t >
void	predict (raft_proto::handle_t const &handle, raft_proto::buffer< io_t > &output, raft_proto::buffer< io_t > const &input, infer_kind predict_type=infer_kind::default_kind, std::optional< index_type > specified_chunk_size=std::nullopt)

template<typename io_t >
void	predict (raft_proto::handle_t const &handle, io_t output, io_t input, std::size_t num_rows, raft_proto::device_type out_mem_type, raft_proto::device_type in_mem_type, infer_kind predict_type=infer_kind::default_kind, std::optional< index_type > specified_chunk_size=std::nullopt)

Detailed Description

A model used for performing inference with FIL

This struct is a wrapper for all variants of decision_forest supported by a standard FIL build.

Constructor & Destructor Documentation

◆ forest_model()

ML::experimental::fil::forest_model::forest_model ( decision_forest_variant && forest = decision_forest_variant{} )

inline

Wrap a decision_forest in a full forest_model object

Member Function Documentation

◆ device_index()

auto ML::experimental::fil::forest_model::device_index ( )

inline

The ID of the device on which this model is loaded

◆ elem_postprocessing()

auto ML::experimental::fil::forest_model::elem_postprocessing ( )

inline

The operation used for postprocessing each element of the output for a single row

◆ has_vector_leaves()

auto ML::experimental::fil::forest_model::has_vector_leaves ( )

inline

Whether or not leaf nodes use vector outputs

◆ is_double_precision()

auto ML::experimental::fil::forest_model::is_double_precision ( )

inline

Whether or not model is loaded at double precision

◆ memory_type()

auto ML::experimental::fil::forest_model::memory_type ( )

inline

The type of memory (device/host) where the model is stored

◆ num_features()

auto ML::experimental::fil::forest_model::num_features ( )

inline

The number of features per row expected by the model

◆ num_outputs()

auto ML::experimental::fil::forest_model::num_outputs ( )

inline

The number of outputs per row generated by the model

◆ num_trees()

auto ML::experimental::fil::forest_model::num_trees ( )

inline

The number of trees in the model

◆ predict() [1/3]

template<typename io_t >

void ML::experimental::fil::forest_model::predict	(	raft_proto::buffer< io_t > &	output,
		raft_proto::buffer< io_t > const &	input,
		raft_proto::cuda_stream	stream = `raft_proto::cuda_stream{}`,
		infer_kind	predict_type = `infer_kind::default_kind`,
		std::optional< index_type >	specified_chunk_size = `std::nullopt`
	)

inline

Perform inference on given input

Parameters

[out]	output	The buffer where model output should be stored. This must be of size at least ROWS x num_outputs().
[in]	input	The buffer containing input data.
[in]	stream	A raft_proto::cuda_stream, which (on GPU-enabled builds) is a transparent wrapper for the cudaStream_t or (on CPU-only builds) a CUDA-free placeholder object.
[in]	predict_type	Type of inference to perform. Defaults to summing the outputs of all trees and produce an output per row. If set to "per_tree", we will instead output all outputs of individual trees. If set to "leaf_id", we will output the integer ID of the leaf node for each tree.
[in]	specified_chunk_size	Specifies the mini-batch size for processing. This has different meanings on CPU and GPU, but on GPU it corresponds to the number of rows evaluated per inference iteration on a single block. It can take on any power of 2 from 1 to 32, and runtime performance is quite sensitive to the value chosen. In general, larger batches benefit from higher values, but it is hard to predict the optimal value a priori. If omitted, a heuristic will be used to select a reasonable value. On CPU, this argument can generally just be omitted.

◆ predict() [2/3]

template<typename io_t >

void ML::experimental::fil::forest_model::predict	(	raft_proto::handle_t const &	handle,
		io_t *	output,
		io_t *	input,
		std::size_t	num_rows,
		raft_proto::device_type	out_mem_type,
		raft_proto::device_type	in_mem_type,
		infer_kind	predict_type = `infer_kind::default_kind`,
		std::optional< index_type >	specified_chunk_size = `std::nullopt`
	)

inline

Perform inference on given input

Parameters

[in]	handle	The raft_proto::handle_t (wrapper for raft::handle_t on GPU) which will be used to provide streams for evaluation.
[out]	output	Pointer to the memory location where output should end up
[in]	input	Pointer to the input data
[in]	num_rows	Number of rows in input
[in]	out_mem_type	The memory type (device/host) of the output buffer
[in]	in_mem_type	The memory type (device/host) of the input buffer
[in]	predict_type	Type of inference to perform. Defaults to summing the outputs of all trees and produce an output per row. If set to "per_tree", we will instead output all outputs of individual trees. If set to "leaf_id", we will output the integer ID of the leaf node for each tree.
[in]	specified_chunk_size	Specifies the mini-batch size for processing. This has different meanings on CPU and GPU, but on GPU it corresponds to the number of rows evaluated per inference iteration on a single block. It can take on any power of 2 from 1 to 32, and runtime performance is quite sensitive to the value chosen. In general, larger batches benefit from higher values, but it is hard to predict the optimal value a priori. If omitted, a heuristic will be used to select a reasonable value. On CPU, this argument can generally just be omitted.

◆ predict() [3/3]

template<typename io_t >

void ML::experimental::fil::forest_model::predict	(	raft_proto::handle_t const &	handle,
		raft_proto::buffer< io_t > &	output,
		raft_proto::buffer< io_t > const &	input,
		infer_kind	predict_type = `infer_kind::default_kind`,
		std::optional< index_type >	specified_chunk_size = `std::nullopt`
	)

inline

Perform inference on given input

Parameters

[in]	handle	The raft_proto::handle_t (wrapper for raft::handle_t on GPU) which will be used to provide streams for evaluation.
[out]	output	The buffer where model output should be stored. If this buffer is on host while the model is on device or vice versa, work will be distributed across available streams to copy the data back to this output location. This must be of size at least ROWS x num_outputs().
[in]	input	The buffer containing input data. If this buffer is on host while the model is on device or vice versa, work will be distributed across available streams to copy the input data to the appropriate location and perform inference.
[in]	predict_type	Type of inference to perform. Defaults to summing the outputs of all trees and produce an output per row. If set to "per_tree", we will instead output all outputs of individual trees. If set to "leaf_id", we will output the integer ID of the leaf node for each tree.
[in]	specified_chunk_size	Specifies the mini-batch size for processing. This has different meanings on CPU and GPU, but on GPU it corresponds to the number of rows evaluated per inference iteration on a single block. It can take on any power of 2 from 1 to 32, and runtime performance is quite sensitive to the value chosen. In general, larger batches benefit from higher values, but it is hard to predict the optimal value a priori. If omitted, a heuristic will be used to select a reasonable value. On CPU, this argument can generally just be omitted.

◆ row_postprocessing()

auto ML::experimental::fil::forest_model::row_postprocessing ( )

inline

The operation used for postprocessing all outputs for a single row

◆ set_row_postprocessing()

void ML::experimental::fil::forest_model::set_row_postprocessing ( row_op val )

inline

Setter for row_postprocessing()

The documentation for this struct was generated from the following file:

include/cuml/experimental/fil/forest_model.hpp

Public Member Functions

Detailed Description

Constructor & Destructor Documentation

◆ forest_model()

Member Function Documentation

◆ device_index()

◆ elem_postprocessing()

◆ has_vector_leaves()

◆ is_double_precision()

◆ memory_type()

◆ num_features()

◆ num_outputs()

◆ num_trees()

◆ predict() [1/3]

◆ predict() [2/3]

◆ predict() [3/3]

◆ row_postprocessing()

◆ set_row_postprocessing()