Struct Model 

pub struct Model { /* private fields */ }

Represents a loaded ML model

Implementations

impl Model

pub fn load<P: AsRef<Path>>(path: P) -> Result<Self>

Load a model from an ONNX file

Example

use ronn_api::Model;

let model = Model::load("model.onnx")?;

pub fn from_bytes(bytes: &[u8]) -> Result<Self>

Load a model from bytes

Example

use ronn_api::Model;

let bytes = std::fs::read("model.onnx")?;
let model = Model::from_bytes(&bytes)?;

pub fn create_session_default(&self) -> Result<InferenceSession>

Create an inference session with default options

Example

use ronn_api::Model;

let model = Model::load("model.onnx")?;
let session = model.create_session_default()?;

pub fn create_session( &self, options: SessionOptions, ) -> Result<InferenceSession>

Create an inference session with custom options

Example

use ronn_api::{Model, SessionOptions, OptimizationLevel};
use ronn_providers::ProviderType;

let model = Model::load("model.onnx")?;
let options = SessionOptions::new()
    .with_optimization_level(OptimizationLevel::O3)
    .with_provider(ProviderType::GPU);
let session = model.create_session(options)?;

pub fn producer_name(&self) -> Option<&str>

Get model metadata

pub fn ir_version(&self) -> i64

Get IR version

pub fn input_names(&self) -> Vec<&str>

Get input names

pub fn output_names(&self) -> Vec<&str>

Get output names