Crate tvm_rt

TVM is a compiler stack for deep learning systems.

This crate provides an idiomatic Rust API for TVM runtime.

The TVM runtime API contains the data structures used by higher-level TVM executors. Specifically it exposes the basic types such as NDArray, as well as the more general object system. The TVM object system enables cross-language interoperability including that of closures for all supported languages including C++, and Python.

Re-exports

pub use object::*;

pub use string::*;

pub use crate::errors::*;

pub use crate::function::Function;

pub use crate::module::Module;

pub use crate::ndarray::NDArray;

Modules

array

device

errors

function

This module provides an idiomatic Rust API for creating and working with TVM functions.

graph_rt

map

module

Provides the Module type and methods for working with runtime TVM modules.

ndarray

This module implements the NDArray type for working with TVM tensors or coverting from a Rust’s ndarray to TVM NDArray.

object

string

Macros

check_call

external

tvm_call

Structs

DataType

Device

Enums

ArgValue

A borrowed TVMPODValue. Can be constructed using into() but the preferred way to obtain a ArgValue is automatically via call_packed!.

ByteArray

A newtype wrapping a raw TVM byte-array.

DeviceType

Device type represents the set of devices supported by TVM.

RetValue

An owned TVMPODValue. Can be converted from a variety of primitive and object types. Can be downcasted using try_from if it contains the desired type.

Functions

get_last_error

Gets the last error message.

version

Outputs the current TVM version.