Quick Start: Embeddings

import dev.langchain4j.model.bedrock.BedrockCohereEmbeddingModel;
import dev.langchain4j.data.segment.TextSegment;
import dev.langchain4j.data.embedding.Embedding;
import dev.langchain4j.model.output.Response;
import software.amazon.awssdk.regions.Region;

BedrockCohereEmbeddingModel model = BedrockCohereEmbeddingModel.builder()
    .region(Region.US_EAST_1)
    .model(BedrockCohereEmbeddingModel.Model.COHERE_EMBED_ENGLISH_V3)
    .inputType(BedrockCohereEmbeddingModel.InputType.SEARCH_DOCUMENT)
    .build();

TextSegment segment = TextSegment.from("Machine learning is a subset of AI");
Response<Embedding> response = model.embed(segment);
float[] vector = response.content().vector();

import dev.langchain4j.model.bedrock.BedrockTitanEmbeddingModel;

BedrockTitanEmbeddingModel model = BedrockTitanEmbeddingModel.builder()
    .region(Region.US_EAST_1)
    .model(BedrockTitanEmbeddingModel.Types.TitanEmbedTextV2.getValue())
    .dimensions(512)
    .normalize(true)
    .build();

Embedding embedding = model.embed("Sample text").content();

import java.util.List;

List<TextSegment> documents = List.of(
    TextSegment.from("Document 1 about ML"),
    TextSegment.from("Document 2 about AI"),
    TextSegment.from("Document 3 about NLP")
);

Response<List<Embedding>> embeddings = model.embedAll(documents);
for (Embedding emb : embeddings.content()) {
    float[] vector = emb.vector();
    // Store or process vector
}

import dev.langchain4j.store.embedding.EmbeddingStore;
import dev.langchain4j.store.embedding.inmemory.InMemoryEmbeddingStore;
import dev.langchain4j.store.embedding.EmbeddingMatch;

// Document model for indexing
BedrockCohereEmbeddingModel docModel = BedrockCohereEmbeddingModel.builder()
    .model(BedrockCohereEmbeddingModel.Model.COHERE_EMBED_ENGLISH_V3)
    .inputType(BedrockCohereEmbeddingModel.InputType.SEARCH_DOCUMENT)
    .build();

// Query model for searching
BedrockCohereEmbeddingModel queryModel = BedrockCohereEmbeddingModel.builder()
    .model(BedrockCohereEmbeddingModel.Model.COHERE_EMBED_ENGLISH_V3)
    .inputType(BedrockCohereEmbeddingModel.InputType.SEARCH_QUERY)
    .build();

// Index documents
EmbeddingStore<TextSegment> store = new InMemoryEmbeddingStore<>();
List<TextSegment> docs = loadDocuments();
Response<List<Embedding>> docEmbeddings = docModel.embedAll(docs);

for (int i = 0; i < docs.size(); i++) {
    store.add(docEmbeddings.content().get(i), docs.get(i));
}

// Search
String query = "neural networks";
Embedding queryEmbedding = queryModel.embed(query).content();
List<EmbeddingMatch<TextSegment>> matches = store.findRelevant(queryEmbedding, 3);

for (EmbeddingMatch<TextSegment> match : matches) {
    System.out.println("Score: " + match.score());
    System.out.println("Text: " + match.embedded().text());
}