Skip to main content

InferenceClient

atomr_agents_agent

Trait InferenceClient 

Source 
pub trait InferenceClient:
    Send
    + Sync
    + 'static {
    // Required methods
    fn provider(&self) -> Provider;
    fn run<'life0, 'async_trait>(
        &'life0 self,
        batch: ExecuteBatch,
    ) -> Pin<Box<dyn Future<Output = Result<TurnResult>> + Send + 'async_trait>>
       where Self: 'async_trait,
             'life0: 'async_trait;
}
Expand description

Implemented by anything the agent can use to drive an inference request. The provider matters for tool-call delta parsing.

Required Methods§

Source

fn provider(&self) -> Provider

Source

fn run<'life0, 'async_trait>( &'life0 self, batch: ExecuteBatch, ) -> Pin<Box<dyn Future<Output = Result<TurnResult>> + Send + 'async_trait>>
where Self: 'async_trait, 'life0: 'async_trait,

Implementors§