echogarden
Version:
An easy-to-use speech toolset. Includes tools for synthesis, recognition, alignment, speech translation, language detection, source separation and more.
13 lines (12 loc) • 745 B
TypeScript
import type * as Onnx from 'onnxruntime-node';
import { OnnxExecutionProvider } from '../utilities/OnnxUtilities.js';
import { RawAudio } from "../audio/AudioUtilities.js";
export declare function computeEmbeddings(audioSamples: RawAudio, modelFilePath: string, executionProviders: OnnxExecutionProvider[]): Promise<Float32Array<ArrayBufferLike>[]>;
export declare class Wav2Vec2BertFeatureEmbeddings {
readonly modelFilePath: string;
readonly executionProviders: OnnxExecutionProvider[];
session?: Onnx.InferenceSession;
constructor(modelFilePath: string, executionProviders: OnnxExecutionProvider[]);
computeEmbeddings(rawAudio: RawAudio): Promise<Float32Array<ArrayBufferLike>[]>;
private initializeSessionIfNeeded;
}