@anvia/transformers: Getting Started

Install

pnpm add @anvia/transformers @anvia/core @huggingface/transformers

Minimum setup

import { embedDocuments } from "@anvia/core/embeddings";
import { InMemoryVectorStore } from "@anvia/core/vector-store";
import { createTransformersEmbeddingModel } from "@anvia/transformers";

const embeddingModel = await createTransformersEmbeddingModel();

const documents = await embedDocuments(
  embeddingModel,
  [{ id: "password-reset", text: "Password reset links expire after 30 minutes." }],
  {
    id: (document) => document.id,
    content: (document) => document.text,
  },
);

const store = InMemoryVectorStore.fromDocuments(documents);
const index = store.index(embeddingModel);

const results = await index.search({
  query: "How long does a reset link last?",
  topK: 3,
});

console.log(results);

Default model

Import DEFAULT_TRANSFORMERS_EMBEDDING_MODEL when the application needs to display or log the default embedding model. Pass an explicit model/options object to createTransformersEmbeddingModel(...) when dimensions, batching, pooling, or runtime behavior need to be controlled.

Next step

Continue with Usage Patterns.