xybrid-core 0.1.0-rc4

Core runtime for hybrid cloud-edge AI inference: model execution, pipeline orchestration, and routing primitives.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

use crate::ir::Envelope;
use crate::runtime_adapter::AdapterResult;
use std::path::Path;

/// Trait for model runtime implementations used by TemplateExecutor.
///
/// This trait abstracts the specific execution logic (Candle, ONNX, TTS, etc.)
/// from the generic TemplateExecutor logic.
pub trait ModelRuntime: Send + Sync {
    /// Returns the name of the runtime (e.g., "candle-whisper", "kokoro-tts")
    fn name(&self) -> &str;

    /// Returns supported file extensions
    fn supported_formats(&self) -> Vec<&str>;

    /// Load the model from the specified path
    fn load(&mut self, model_path: &Path) -> AdapterResult<()>;

    /// Downcast to concrete type
    fn as_any(&self) -> &dyn std::any::Any;

    /// Execute inference on the input envelope
    fn execute(&mut self, input: &Envelope) -> AdapterResult<Envelope>;
}