获取文本嵌入

本文档介绍如何使用 Vertex AI 文本嵌入 API 创建文本嵌入。

Vertex AI 文本嵌入 API 使用密集向量表示法:例如,text-embedding-gecko 使用 768 维向量。密集向量嵌入模型使用与大语言模型所用方法类似的深度学习方法。与倾向于将字词直接映射到数字的稀疏向量不同,密集向量旨在更好地表示一段文本的含义。在生成式 AI 中使用密集向量嵌入的优势在于,您可以更好地搜索与查询含义相符的段落,而不是搜索直接的字词或语法匹配项,即使段落不使用相同的语言也是如此。

这些向量已进行标准化处理,因此您可以使用余弦相似度、点积或欧几里得距离来提供相同的相似度排名。

准备工作

  1. Sign in to your Google Cloud account. If you're new to Google Cloud, create an account to evaluate how our products perform in real-world scenarios. New customers also get $300 in free credits to run, test, and deploy workloads.
  2. In the Google Cloud console, on the project selector page, select or create a Google Cloud project.

    Go to project selector

  3. Enable the Vertex AI API.

    Enable the API

  4. In the Google Cloud console, on the project selector page, select or create a Google Cloud project.

    Go to project selector

  5. Enable the Vertex AI API.

    Enable the API

  6. 为嵌入作业选择任务类型

支持的模型

您可以使用以下模型获取文本嵌入:

英语模型 多语言模型
textembedding-gecko@001 textembedding-gecko-multilingual@001
textembedding-gecko@002 text-multilingual-embedding-002
textembedding-gecko@003
text-embedding-004
text-embedding-005

如果您刚开始接触这些模型,我们建议您使用最新版本。对于英语文本,请使用 text-embedding-005。对于多语言文本,请使用 text-multilingual-embedding-002

获取文本片段的文本嵌入

您可以使用 Vertex AI API 或 Python 版 Vertex AI SDK 获取文本片段的文本嵌入。对于每个请求,在 us-central1,输入文本上限为 250 个,而在其他区域,输入文本数上限为 5。 API 的输入词元数量上限为 20,000。超出此限制的输入将导致 500 错误。每个输入文本进一步限制为 2048 个词元;任何多余的内容都会以静默方式截断。您还可以通过将 autoTruncate 设置为 false 来停用静默截断。

默认情况下,所有模型都会生成具有 768 个维度的输出。不过,以下模型可让用户选择 1 到 768 之间的输出维数。通过选择较小的输出维度,用户可以节省内存和存储空间,从而实现更高效的计算。

  • text-embedding-005
  • text-multilingual-embedding-002

以下示例使用 text-embedding-005 模型。

Gen AI SDK for Python

了解如何安装或更新 Google Gen AI SDK for Python
如需了解详情,请参阅 Gen AI SDK for Python API 参考文档python-genai GitHub 代码库
设置环境变量以将 Gen AI SDK 与 Vertex AI 搭配使用:

# Replace the `GOOGLE_CLOUD_PROJECT` and `GOOGLE_CLOUD_LOCATION` values
# with appropriate values for your project.
export GOOGLE_CLOUD_PROJECT=GOOGLE_CLOUD_PROJECT
export GOOGLE_CLOUD_LOCATION=us-central1
export GOOGLE_GENAI_USE_VERTEXAI=True

from google import genai
from google.genai.types import EmbedContentConfig

client = genai.Client()
response = client.models.embed_content(
    model="text-embedding-005",
    contents=[
        "How do I get a driver's license/learner's permit?",
        "How do I renew my driver's license?",
        "How do I change my address on my driver's license?",
    ],
    config=EmbedContentConfig(
        task_type="RETRIEVAL_DOCUMENT",  # Optional
        output_dimensionality=768,  # Optional
        title="Driver's License",  # Optional
    ),
)
print(response)
# Example response:
# embeddings=[ContentEmbedding(values=[-0.06302902102470398, 0.00928034819662571, 0.014716853387653828, -0.028747491538524628, ... ],
# statistics=ContentEmbeddingStatistics(truncated=False, token_count=13.0))]
# metadata=EmbedContentMetadata(billable_character_count=112)

Python 版 Vertex AI SDK

如需了解如何安装或更新 Vertex AI SDK for Python,请参阅安装 Vertex AI SDK for Python。 如需了解详情,请参阅 Python 版 Vertex AI SDK API 参考文档

from __future__ import annotations

from vertexai.language_models import TextEmbeddingInput, TextEmbeddingModel


def embed_text() -> list[list[float]]:
    """Embeds texts with a pre-trained, foundational model.

    Returns:
        A list of lists containing the embedding vectors for each input text
    """

    # A list of texts to be embedded.
    texts = ["banana muffins? ", "banana bread? banana muffins?"]
    # The dimensionality of the output embeddings.
    dimensionality = 256
    # The task type for embedding. Check the available tasks in the model's documentation.
    task = "RETRIEVAL_DOCUMENT"

    model = TextEmbeddingModel.from_pretrained("text-embedding-005")
    inputs = [TextEmbeddingInput(text, task) for text in texts]
    kwargs = dict(output_dimensionality=dimensionality) if dimensionality else {}
    embeddings = model.get_embeddings(inputs, **kwargs)

    print(embeddings)
    # Example response:
    # [[0.006135190837085247, -0.01462465338408947, 0.004978656303137541, ...], [0.1234434666, ...]],
    return [embedding.values for embedding in embeddings]

Go

在尝试此示例之前,请按照《Vertex AI 快速入门:使用客户端库》中的 Go 设置说明执行操作。 如需了解详情,请参阅 Vertex AI Go API 参考文档

如需向 Vertex AI 进行身份验证,请设置应用默认凭据。 如需了解详情,请参阅为本地开发环境设置身份验证

import (
	"context"
	"fmt"
	"io"

	aiplatform "cloud.google.com/go/aiplatform/apiv1"
	"cloud.google.com/go/aiplatform/apiv1/aiplatformpb"

	"google.golang.org/api/option"
	"google.golang.org/protobuf/types/known/structpb"
)

// embedTexts shows how embeddings are set for text-embedding-005 model
func embedTexts(w io.Writer, project, location string) error {
	// location := "us-central1"
	ctx := context.Background()

	apiEndpoint := fmt.Sprintf("%s-aiplatform.googleapis.com:443", location)
	dimensionality := 5
	model := "text-embedding-005"
	texts := []string{"banana muffins? ", "banana bread? banana muffins?"}

	client, err := aiplatform.NewPredictionClient(ctx, option.WithEndpoint(apiEndpoint))
	if err != nil {
		return err
	}
	defer client.Close()

	endpoint := fmt.Sprintf("projects/%s/locations/%s/publishers/google/models/%s", project, location, model)
	instances := make([]*structpb.Value, len(texts))
	for i, text := range texts {
		instances[i] = structpb.NewStructValue(&structpb.Struct{
			Fields: map[string]*structpb.Value{
				"content":   structpb.NewStringValue(text),
				"task_type": structpb.NewStringValue("QUESTION_ANSWERING"),
			},
		})
	}

	params := structpb.NewStructValue(&structpb.Struct{
		Fields: map[string]*structpb.Value{
			"outputDimensionality": structpb.NewNumberValue(float64(dimensionality)),
		},
	})

	req := &aiplatformpb.PredictRequest{
		Endpoint:   endpoint,
		Instances:  instances,
		Parameters: params,
	}
	resp, err := client.Predict(ctx, req)
	if err != nil {
		return err
	}
	embeddings := make([][]float32, len(resp.Predictions))
	for i, prediction := range resp.Predictions {
		values := prediction.GetStructValue().Fields["embeddings"].GetStructValue().Fields["values"].GetListValue().Values
		embeddings[i] = make([]float32, len(values))
		for j, value := range values {
			embeddings[i][j] = float32(value.GetNumberValue())
		}
	}

	fmt.Fprintf(w, "Dimensionality: %d. Embeddings length: %d", len(embeddings[0]), len(embeddings))
	return nil
}

Java

在尝试此示例之前,请按照《Vertex AI 快速入门:使用客户端库》中的 Java 设置说明执行操作。 如需了解详情,请参阅 Vertex AI Java API 参考文档

如需向 Vertex AI 进行身份验证,请设置应用默认凭据。 如需了解详情,请参阅为本地开发环境设置身份验证

import static java.util.stream.Collectors.toList;

import com.google.cloud.aiplatform.v1.EndpointName;
import com.google.cloud.aiplatform.v1.PredictRequest;
import com.google.cloud.aiplatform.v1.PredictResponse;
import com.google.cloud.aiplatform.v1.PredictionServiceClient;
import com.google.cloud.aiplatform.v1.PredictionServiceSettings;
import com.google.protobuf.Struct;
import com.google.protobuf.Value;
import java.io.IOException;
import java.util.ArrayList;
import java.util.List;
import java.util.OptionalInt;
import java.util.regex.Matcher;
import java.util.regex.Pattern;

public class PredictTextEmbeddingsSample {
  public static void main(String[] args) throws IOException {
    // TODO(developer): Replace these variables before running the sample.
    // Details about text embedding request structure and supported models are available in:
    // https://cloud.google.com/vertex-ai/docs/generative-ai/embeddings/get-text-embeddings
    String endpoint = "us-central1-aiplatform.googleapis.com:443";
    String project = "YOUR_PROJECT_ID";
    String model = "text-embedding-005";
    predictTextEmbeddings(
        endpoint,
        project,
        model,
        List.of("banana bread?", "banana muffins?"),
        "QUESTION_ANSWERING",
        OptionalInt.of(256));
  }

  // Gets text embeddings from a pretrained, foundational model.
  public static List<List<Float>> predictTextEmbeddings(
      String endpoint,
      String project,
      String model,
      List<String> texts,
      String task,
      OptionalInt outputDimensionality)
      throws IOException {
    PredictionServiceSettings settings =
        PredictionServiceSettings.newBuilder().setEndpoint(endpoint).build();
    Matcher matcher = Pattern.compile("^(?<Location>\\w+-\\w+)").matcher(endpoint);
    String location = matcher.matches() ? matcher.group("Location") : "us-central1";
    EndpointName endpointName =
        EndpointName.ofProjectLocationPublisherModelName(project, location, "google", model);

    // You can use this prediction service client for multiple requests.
    try (PredictionServiceClient client = PredictionServiceClient.create(settings)) {
      PredictRequest.Builder request =
          PredictRequest.newBuilder().setEndpoint(endpointName.toString());
      if (outputDimensionality.isPresent()) {
        request.setParameters(
            Value.newBuilder()
                .setStructValue(
                    Struct.newBuilder()
                        .putFields("outputDimensionality", valueOf(outputDimensionality.getAsInt()))
                        .build()));
      }
      for (int i = 0; i < texts.size(); i++) {
        request.addInstances(
            Value.newBuilder()
                .setStructValue(
                    Struct.newBuilder()
                        .putFields("content", valueOf(texts.get(i)))
                        .putFields("task_type", valueOf(task))
                        .build()));
      }
      PredictResponse response = client.predict(request.build());
      List<List<Float>> floats = new ArrayList<>();
      for (Value prediction : response.getPredictionsList()) {
        Value embeddings = prediction.getStructValue().getFieldsOrThrow("embeddings");
        Value values = embeddings.getStructValue().getFieldsOrThrow("values");
        floats.add(
            values.getListValue().getValuesList().stream()
                .map(Value::getNumberValue)
                .map(Double::floatValue)
                .collect(toList()));
      }
      return floats;
    }
  }

  private static Value valueOf(String s) {
    return Value.newBuilder().setStringValue(s).build();
  }

  private static Value valueOf(int n) {
    return Value.newBuilder().setNumberValue(n).build();
  }
}

Node.js

在尝试此示例之前,请按照《Vertex AI 快速入门:使用客户端库》中的 Node.js 设置说明执行操作。 如需了解详情,请参阅 Vertex AI Node.js API 参考文档

如需向 Vertex AI 进行身份验证,请设置应用默认凭据。 如需了解详情,请参阅为本地开发环境设置身份验证

async function main(
  project,
  model = 'text-embedding-005',
  texts = 'banana bread?;banana muffins?',
  task = 'QUESTION_ANSWERING',
  dimensionality = 0,
  apiEndpoint = 'us-central1-aiplatform.googleapis.com'
) {
  const aiplatform = require('@google-cloud/aiplatform');
  const {PredictionServiceClient} = aiplatform.v1;
  const {helpers} = aiplatform; // helps construct protobuf.Value objects.
  const clientOptions = {apiEndpoint: apiEndpoint};
  const location = 'us-central1';
  const endpoint = `projects/${project}/locations/${location}/publishers/google/models/${model}`;

  async function callPredict() {
    const instances = texts
      .split(';')
      .map(e => helpers.toValue({content: e, task_type: task}));
    const parameters = helpers.toValue(
      dimensionality > 0 ? {outputDimensionality: parseInt(dimensionality)} : {}
    );
    const request = {endpoint, instances, parameters};
    const client = new PredictionServiceClient(clientOptions);
    const [response] = await client.predict(request);
    const predictions = response.predictions;
    const embeddings = predictions.map(p => {
      const embeddingsProto = p.structValue.fields.embeddings;
      const valuesProto = embeddingsProto.structValue.fields.values;
      return valuesProto.listValue.values.map(v => v.numberValue);
    });
    console.log('Got embeddings: \n' + JSON.stringify(embeddings));
  }

  callPredict();
}

将嵌入添加到向量数据库

生成嵌入后,您可以将嵌入添加到向量数据库,例如 Vector Search。这样可以实现低延迟检索,并且随着数据规模扩大,这一点至关重要。

如需详细了解 Vector Search,请参阅 Vector Search 概览

后续步骤