@huggingface/inference/dist/commonjs/tasks/nlp/tokenClassification.js

Typescript client for the Hugging Face Inference Providers and Inference Endpoints

"use strict";
Object.defineProperty(exports, "__esModule", { value: true });
exports.tokenClassification = tokenClassification;
const getInferenceProviderMapping_js_1 = require("../../lib/getInferenceProviderMapping.js");
const getProviderHelper_js_1 = require("../../lib/getProviderHelper.js");
const request_js_1 = require("../../utils/request.js");
/**
 * Usually used for sentence parsing, either grammatical, or Named Entity Recognition (NER) to understand keywords contained within text. Recommended model: dbmdz/bert-large-cased-finetuned-conll03-english
 */
async function tokenClassification(args, options) {
    const provider = await (0, getInferenceProviderMapping_js_1.resolveProvider)(args.provider, args.model, args.endpointUrl);
    const providerHelper = (0, getProviderHelper_js_1.getProviderHelper)(provider, "token-classification");
    const { data: res } = await (0, request_js_1.innerRequest)(args, providerHelper, {
        ...options,
        task: "token-classification",
    });
    return providerHelper.getResponse(res);
}