Type Alias: WhisperSttModel<L>

WhisperSttModel<L> = object

Defined in: extensions/speech/tasks/whisperSpeechToText.ts:93

Paths and metadata required to instantiate a Whisper speech-to-text model.

Type Parameters

L

L extends WhisperLanguage = WhisperLanguage

Properties

modelPath

readonly modelPath: string

Defined in: extensions/speech/tasks/whisperSpeechToText.ts:97

Local path or remote URL of the .pte model.

---

name?

readonly optional name: "whisper"

Defined in: extensions/speech/tasks/whisperSpeechToText.ts:95

Pipeline tag, which tells useSpeechToText which runner to build.

---

supportedLanguages

readonly supportedLanguages: readonly L[]

Defined in: extensions/speech/tasks/whisperSpeechToText.ts:101

List of supported language codes for this model.

---

tokenizerPath

readonly tokenizerPath: string

Defined in: extensions/speech/tasks/whisperSpeechToText.ts:99

Local path or remote URL of the tokenizer file.

---

vadModel

readonly vadModel: FsmnVadModel

Defined in: extensions/speech/tasks/whisperSpeechToText.ts:103

VAD model configuration used for speech segmentation.