README.md

metadata

license: mit
base_model:
  - sentence-transformers/msmarco-distilbert-cos-v5

Usage using Transformer.js for feature extraction

import { env, pipeline } from "@xenova/transformers";

env.allowLocalModels = false;

const extractor = await pipeline(
  "feature-extraction",
  "Fuzail22/onnx-msmarco-distilbert-cos-v5",
  {
    quantized: false
  }
);

const extract = async (text: string) => {
    const output = await extractor(text, {
      pooling: "mean",
      normalize: true
    });
    console.log("extraction output:", output);
};

extact("Open Source is great!")

This is Onnx converted version of the original model The Input Sequence length is 512

Converted Using Xenova

Link to original Model sentence-transformers/msmarco-distilbert-cos-v5