Struct TranscriptionConfig 

pub struct TranscriptionConfig {
    pub language: Option<String>,
    pub beam_size: usize,
    pub word_timestamps: bool,
    pub write_sidecar: bool,
    pub backend: TranscriptionBackend,
    pub model_path: Option<PathBuf>,
    pub prompt: Option<String>,
    pub hotwords: Vec<String>,
}

Configuration for the transcription pipeline.

Fields

language: Option<String>

Language hint (ISO 639-1, e.g., “en”). None for auto-detect.

beam_size: usize

Beam size for decoding (1 = greedy, 5 = default).

word_timestamps: bool

Enable word-level timestamps (more precise but slower).

write_sidecar: bool

Write .srt sidecar files after transcription for caching.

backend: TranscriptionBackend

Compute backend for inference.

model_path: Option<PathBuf>

Path to the .apr model file (e.g. base.apr). When None, transcription of files without sidecars will fail with a helpful error message.

prompt: Option<String>

Initial prompt to condition the decoder on domain vocabulary. Example: “This lecture covers AWS, Kubernetes, and YAML configurations.”

hotwords: Vec<String>

Hotwords to boost during decoding for domain-specific terms. Each string is a word or phrase to bias positively in the logit space.

Trait Implementations

impl Clone for TranscriptionConfig

fn clone(&self) -> TranscriptionConfig

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for TranscriptionConfig

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for TranscriptionConfig

fn default() -> Self

Returns the “default value” for a type.