pub struct ModelFacts {
pub context_tokens: Option<u64>,
pub parameters_b: Option<f64>,
pub served_locally: bool,
}Expand description
What a backend was able to tell us about a model and how it is served.
Both measurements are independently optional: no backend reports both, and
a bare llama-server reports neither. served_locally is always known, and
is what decides the case where the measurements are absent.
Fields§
§context_tokens: Option<u64>Usable context window in tokens. Ollama reports it via /api/show,
OpenRouter via context_length; a bare llama-server does not.
parameters_b: Option<f64>Parameter count in billions. Ollama reports it; hosted providers generally do not.
served_locally: boolWhether the endpoint is a model served on this machine or the local network. Decides the profile when neither measurement is available: what people run locally is small and configured modestly, and the observed failure there is a hard refusal rather than a slightly narrower run.
Trait Implementations§
Source§impl Clone for ModelFacts
impl Clone for ModelFacts
Source§fn clone(&self) -> ModelFacts
fn clone(&self) -> ModelFacts
1.0.0 (const: unstable) · Source§fn clone_from(&mut self, source: &Self)
fn clone_from(&mut self, source: &Self)
source. Read more