import { pipeline, cos_sim } from '@huggingface/transformers';

// Create a feature extraction pipeline
const extractor = await pipeline('feature-extraction', 'Xenova/jina-embeddings-v2-base-zh', {
    dtype: "fp32"  // Options: "fp32", "fp16", "q8", "q4"
});

// Compute sentence embeddings
const texts = ['How is the weather today?', '今天天气怎么样?'];
const output = await extractor(texts, { pooling: 'mean', normalize: true });
// Tensor {
//   dims: [2, 768],
// 	 type: 'float32',
//   data: Float32Array(1536)[...],
// 	 size: 1536
// }

// Compute cosine similarity between the two embeddings
const score = cos_sim(output[0].data, output[1].data);
console.log(score);
// 0.7860610759096025

Note: Having a separate repo for ONNX weights is intended to be a temporary solution until WebML gains more traction. If you would like to make your models web-ready, we recommend converting to ONNX using 🤗 Optimum and structuring your repo like this one (with ONNX weights located in a subfolder named onnx).