Grounding

Dalam AI generatif, perujukan adalah kemampuan untuk menghubungkan output model ke sumber informasi yang dapat diverifikasi. Jika Anda memberi model akses ke sumber data tertentu, grounding akan mengikat outputnya ke data ini dan mengurangi kemungkinan pembuatan konten.

Dengan Vertex AI, Anda dapat mendasarkan output model dengan cara berikut:

  • Grounding dengan Google Penelusuran - melakukan grounding pada model dengan data web yang tersedia secara publik.
  • Ground to your own data - membuat model dengan data Anda sendiri dari Vertex AI Search sebagai penyimpanan data (Pratinjau).

Untuk informasi selengkapnya tentang pentanahan, lihat Ringkasan pentanahan.

Model yang didukung

Model Versi
Gemini 2.0 Flash hanya dengan input teks gemini-2.0-flash-001
Gemini 1.5 Pro hanya dengan input teks gemini-1.5-pro-002
gemini-1.5-pro-001
Gemini 1.5 Flash hanya dengan input teks gemini-1.5-flash-002
gemini-1.5-flash-001
Gemini 1.0 Pro hanya dengan input teks gemini-1.0-pro-001
gemini-1.0-pro-002

Batasan

Pembumian hanya tersedia untuk permintaan teks.

Contoh sintaksis

Contoh ini menunjukkan sintaksis untuk cara melandasi model dan menentukan bahasa yang didukung. Untuk dukungan bahasa, lihat Bahasa.


Tentukan nilai untuk variabel berikut:
  • PROMPT_TEXT: Permintaan teks yang Anda kirimkan ke model bahasa untuk menerima respons balik.
  • SYSTEM_INSTRUCTION: Kumpulan petunjuk yang memberi tahu model cara berperilaku dan merespons perintah.
  • FACT_TEXT_1: Informasi terverifikasi yang digunakan oleh model untuk menghasilkan respons.
  • TITLE_1: Nama sumber informasi terverifikasi Anda.
  • URI_1: String yang secara unik mengidentifikasi lokasi sumber informasi terverifikasi Anda.
  • AUTHOR_1: Penyedia informasi terverifikasi yang digunakan oleh model untuk menghasilkan respons.
  • FACT_TEXT_2: Informasi terverifikasi yang digunakan oleh model untuk menghasilkan respons.
  • TITLE_2: Nama sumber informasi terverifikasi Anda.
  • URI_2: String yang secara unik mengidentifikasi lokasi sumber informasi terverifikasi Anda.
  • FACT_TEXT_3: Informasi terverifikasi yang digunakan oleh model untuk menghasilkan respons.
  • TITLE_3: Nama sumber informasi terverifikasi Anda.
  • URI_3: String yang secara unik mengidentifikasi lokasi sumber informasi terverifikasi Anda.
  • PROJECT_NUMBER: Mengidentifikasi Google Cloud project Anda secara unik, yang berisi resource Anda.
  • APP_ID_1: Aplikasi yang berinteraksi dengan model.
  • APP_ID_2: Aplikasi yang berinteraksi dengan model.
  • MODEL_ID: Mengidentifikasi model yang berinteraksi dengan aplikasi Anda.
  • LANGUAGE_CODE: Mengidentifikasi bahasa yang digunakan dalam membuat respons yang beralasan.

curl

curl -X POST \
-H "Authorization: Bearer $(gcloud auth print-access-token)" \
-H "Content-Type: application/json" \
"https://discoveryengine.googleapis.com/v1/projects/PROJECT_NUMBER/locations/global:generateGroundedContent" \
-d '
{
"contents": [
 {
   "role": "user",
   "parts": [
     {
       "text": "PROMPT_TEXT"
     }
   ]
 }
],
"systemInstruction": {
   "parts": {
       "text": "SYSTEM_INSTRUCTION"
   }
},
"groundingSpec": {
 "groundingSources": [
   {
     "inlineSource": {
       "groundingFacts": [
         {
           "factText": "FACT_TEXT_1",
           "attributes": {
             "title": "TITLE_1",
             "uri": "URI_1",
             "author": "AUTHOR_1"
           }
         }
       ]
     }
   },
   {
     "inlineSource": {
       "groundingFacts": [
         {
           "factText": "FACT_TEXT_2",
           "attributes": {
             "title": "TITLE_2",
             "uri": "URI_2"
           }
         },
         {
           "factText": "FACT_TEXT_3",
           "attributes": {
             "title": "TITLE_3",
             "uri": "URI_3"
           }
         }
       ]
     }
   },
   {
     "searchSource": {
       "servingConfig": "projects/PROJECT_NUMBER/locations/global/collections/default_collection/engines/APP_ID_1/servingConfigs/default_search"
     }
   },
   {
     "searchSource": {
       "servingConfig": "projects/PROJECT_NUMBER/locations/global/collections/default_collection/engines/APP_ID_2/servingConfigs/default_search"
     }
   }
  ]
},
"generationSpec": {
  "modelId": "MODEL_ID",
  "temperature": TEMPERATURE,
  "topP": TOP_P,
  "topK": TOP_K,
},
"user_context": {
  "languageCode: "LANGUAGE_CODE"
  "latLng": {
    "latitude": 46.7,
    "longitude": 8.9
  },
}'

Daftar parameter

Lihat contoh untuk mengetahui detail penerapan.

GoogleSearchRetrieval

Dasari respons dengan data publik.

Parameter

google_search_retrieval

Wajib: Object

Latar belakang dengan data web yang tersedia secara publik.

Retrieval

Dasari respons dengan data pribadi dari Vertex AI Search sebagai penyimpanan data. Menentukan alat pengambilan yang dapat dipanggil model untuk mengakses pengetahuan eksternal.

Parameter

source

Wajib: VertexAISearch

Lakukan dasar-dasar dengan sumber data Vertex AI Search.

VertexAISearch

Parameter

datastore

Wajib: string

ID resource penyimpanan data yang sepenuhnya memenuhi syarat dari Vertex AI Search, dalam format berikut: projects/{project}/locations/{location}/collections/default_collection/dataStores/{datastore}

Contoh

Menghubungkan respons pada data web publik menggunakan Google Penelusuran

Hubungkan respons dengan data publik Google Penelusuran. Sertakan alat google_search_retrieval dalam permintaan. Tidak diperlukan parameter tambahan.

REST

Sebelum menggunakan salah satu data permintaan, lakukan penggantian berikut:

  • LOCATION: Region untuk memproses permintaan.
  • PROJECT_ID: Project ID Anda.
  • MODEL_ID: ID model model multimodal. Hanya model Gemini 1.0 dan Gemini 1.5 yang mendukung pengambilan dinamis. Model Gemini 2.0 tidak mendukung pengambilan dinamis.
  • TEXT: Petunjuk teks yang akan disertakan dalam perintah.
  • DYNAMIC_THRESHOLD: Kolom opsional untuk menetapkan nilai minimum guna memanggil konfigurasi pengambilan dinamis. Ini adalah nilai floating point dalam rentang [0,1]. Jika Anda tidak menetapkan kolom dynamicThreshold, nilai nilai minimum akan ditetapkan secara default ke 0,7.

Metode HTTP dan URL: 

POST https://LOCATION-aiplatform.googleapis.com/v1beta1/projects/PROJECT_ID/locations/LOCATION/publishers/google/models/MODEL_ID:generateContent

Meminta isi JSON: 

{
  "contents": [{
    "role": "user",
    "parts": [{
      "text": "TEXT"
    }]
  }],
  "tools": [{
    "googleSearchRetrieval": {
      "dynamicRetrievalConfig": {
        "mode": "MODE_DYNAMIC",
        "dynamicThreshold": DYNAMIC_THRESHOLD
      }
    }
  }],
  "model": "projects/PROJECT_ID/locations/LOCATION/publishers/google/models/MODEL_ID"
}

Untuk mengirim permintaan Anda, perluas salah satu opsi berikut:

Anda akan melihat respons JSON seperti berikut:

{
   "candidates": [
     {
       "content": {
         "role": "model",
         "parts": [
           {
             "text": "Chicago weather changes rapidly, so layers let you adjust easily. Consider a base layer, a warm mid-layer (sweater-fleece), and a weatherproof outer layer."
           }
         ]
       },
       "finishReason": "STOP",
       "safetyRatings":[
       "..."
    ],
       "groundingMetadata": {
         "webSearchQueries": [
           "What's the weather in Chicago this weekend?"
         ],
         "searchEntryPoint": {
            "renderedContent": "....................."
         }
         "groundingSupports": [
            {
              "segment": {
                "startIndex": 0,
                "endIndex": 65,
                "text": "Chicago weather changes rapidly, so layers let you adjust easily."
              },
              "groundingChunkIndices": [
                0
              ],
              "confidenceScores": [
                0.99
              ]
            },
          ]
          "retrievalMetadata": {
              "webDynamicRetrievalScore": 0.96879
            }
       }
     }
   ],
   "usageMetadata": { "..."
   }
 }

Python 

import vertexai

from vertexai.generative_models import (
    GenerationConfig,
    GenerativeModel,
    Tool,
    grounding,
)

# TODO(developer): Update and un-comment below line
# PROJECT_ID = "your-project-id"
vertexai.init(project=PROJECT_ID, location="us-central1")

model = GenerativeModel("gemini-1.5-flash-001")

# Use Google Search for grounding
tool = Tool.from_google_search_retrieval(
    grounding.GoogleSearchRetrieval(
        # Optional: For Dynamic Retrieval
        dynamic_retrieval_config=grounding.DynamicRetrievalConfig(
            dynamic_threshold=0.7,
        )
    )
)

prompt = "When is the next total solar eclipse in US?"
response = model.generate_content(
    prompt,
    tools=[tool],
    generation_config=GenerationConfig(
        temperature=0.0,
    ),
)

print(response.text)
# Example response:
# The next total solar eclipse visible from the contiguous United States will be on **August 23, 2044**.

NodeJS 

const {VertexAI} = require('@google-cloud/vertexai');

/**
 * TODO(developer): Update these variables before running the sample.
 */
async function generateContentWithGoogleSearchGrounding(
  projectId = 'PROJECT_ID',
  location = 'us-central1',
  model = 'gemini-1.5-flash-001'
) {
  // Initialize Vertex with your Cloud project and location
  const vertexAI = new VertexAI({project: projectId, location: location});

  const generativeModelPreview = vertexAI.preview.getGenerativeModel({
    model: model,
    generationConfig: {maxOutputTokens: 256},
  });

  const googleSearchRetrievalTool = {
    googleSearchRetrieval: {},
  };

  const request = {
    contents: [{role: 'user', parts: [{text: 'Why is the sky blue?'}]}],
    tools: [googleSearchRetrievalTool],
  };

  const result = await generativeModelPreview.generateContent(request);
  const response = await result.response;
  const groundingMetadata = response.candidates[0].groundingMetadata;
  console.log(
    'Response: ',
    JSON.stringify(response.candidates[0].content.parts[0].text)
  );
  console.log('GroundingMetadata is: ', JSON.stringify(groundingMetadata));
}

Java 

import com.google.cloud.vertexai.VertexAI;
import com.google.cloud.vertexai.api.GenerateContentResponse;
import com.google.cloud.vertexai.api.GoogleSearchRetrieval;
import com.google.cloud.vertexai.api.GroundingMetadata;
import com.google.cloud.vertexai.api.Tool;
import com.google.cloud.vertexai.generativeai.GenerativeModel;
import com.google.cloud.vertexai.generativeai.ResponseHandler;
import java.io.IOException;
import java.util.Collections;

public class GroundingWithPublicData {
  public static void main(String[] args) throws IOException {
    // TODO(developer): Replace these variables before running the sample.
    String projectId = "your-google-cloud-project-id";
    String location = "us-central1";
    String modelName = "gemini-1.5-flash-001";

    groundWithPublicData(projectId, location, modelName);
  }

  // A request whose response will be "grounded" with information found in Google Search.
  public static String groundWithPublicData(String projectId, String location, String modelName)
      throws IOException {
    // Initialize client that will be used to send requests.
    // This client only needs to be created once, and can be reused for multiple requests.
    try (VertexAI vertexAI = new VertexAI(projectId, location)) {
      Tool googleSearchTool =
          Tool.newBuilder()
              .setGoogleSearchRetrieval(
                  // Enable using the result from this tool in detecting grounding
                  GoogleSearchRetrieval.newBuilder())
              .build();

      GenerativeModel model =
          new GenerativeModel(modelName, vertexAI)
              .withTools(Collections.singletonList(googleSearchTool));

      GenerateContentResponse response = model.generateContent("Why is the sky blue?");

      GroundingMetadata groundingMetadata = response.getCandidates(0).getGroundingMetadata();
      String answer = ResponseHandler.getText(response);

      System.out.println("Answer: " + answer);
      System.out.println("Grounding metadata: " + groundingMetadata);

      return answer;
    }
  }
}

Respons dasar pada data pribadi menggunakan Vertex AI Search

Dasari respons dengan data dari penyimpanan data Vertex AI Search. Untuk mengetahui informasi selengkapnya, lihat Vertex AI Agent Builder.

Sebelum Anda mendasarkan respons dengan data pribadi, buat penyimpanan data dan aplikasi penelusuran.

PERINGATAN: Untuk saat ini, antarmuka "grounding" ini tidak mendukung "mode bagian" Penelusuran Vertex AI.

REST

Sebelum menggunakan salah satu data permintaan, lakukan penggantian berikut:

  • LOCATION: Region untuk memproses permintaan.
  • PROJECT_ID: Project ID Anda.
  • MODEL_ID: ID model model multimodal.
  • TEXT: Petunjuk teks yang akan disertakan dalam perintah.

Metode HTTP dan URL: 

POST https://LOCATION-aiplatform.googleapis.com/v1beta1/projects/PROJECT_ID/locations/LOCATION/publishers/google/models/MODEL_ID:generateContent

Meminta isi JSON: 

{
  "contents": [{
    "role": "user",
    "parts": [{
      "text": "TEXT"
    }]
  }],
  "tools": [{
    "retrieval": {
      "vertexAiSearch": {
        "datastore": projects/PROJECT_ID/locations/global/collections/default_collection/dataStores/DATA_STORE_ID
      }
    }
  }],
  "model": "projects/PROJECT_ID/locations/LOCATION/publishers/google/models/MODEL_ID"
}

Untuk mengirim permintaan Anda, perluas salah satu opsi berikut:

Anda akan melihat respons JSON seperti berikut:

{
  "candidates": [
    {
      "content": {
        "role": "model",
        "parts": [
          {
            "text": "You can make an appointment on the website https://dmv.gov/"
          }
        ]
      },
      "finishReason": "STOP",
      "safetyRatings": [
        "..."
      ],
      "groundingMetadata": {
        "retrievalQueries": [
          "How to make appointment to renew driving license?"
        ],
        "groundingChunks": [
          {
            "retrievedContext": {
              "uri": "https://vertexaisearch.cloud.google.com/grounding-api-redirect/AXiHM.....QTN92V5ePQ==",
              "title": "dmv"
            }
          }
        ],
        "groundingSupport": [
          {
            "segment": {
              "startIndex": 25,
              "endIndex": 147
            },
            "segment_text": "ipsum lorem ...",
            "supportChunkIndices": [1, 2],
            "confidenceScore": [0.9541752, 0.97726375]
          },
          {
            "segment": {
              "startIndex": 294,
              "endIndex": 439
            },
            "segment_text": "ipsum lorem ...",
            "supportChunkIndices": [1],
            "confidenceScore": [0.9541752, 0.9325467]
          }
        ]
      }
    }
  ],
  "usageMetadata": {
    "..."
  }
}

Python 

import vertexai

from vertexai.preview.generative_models import (
    GenerationConfig,
    GenerativeModel,
    Tool,
    grounding,
)

# TODO(developer): Update and un-comment below lines
# PROJECT_ID = "your-project-id"
# data_store_id = "your-data-store-id"

vertexai.init(project=PROJECT_ID, location="us-central1")

model = GenerativeModel("gemini-1.5-flash-001")

tool = Tool.from_retrieval(
    grounding.Retrieval(
        grounding.VertexAISearch(
            datastore=data_store_id,
            project=PROJECT_ID,
            location="global",
        )
    )
)

prompt = "How do I make an appointment to renew my driver's license?"
response = model.generate_content(
    prompt,
    tools=[tool],
    generation_config=GenerationConfig(
        temperature=0.0,
    ),
)

print(response.text)

NodeJS 

const {
  VertexAI,
  HarmCategory,
  HarmBlockThreshold,
} = require('@google-cloud/vertexai');

/**
 * TODO(developer): Update these variables before running the sample.
 */
async function generateContentWithVertexAISearchGrounding(
  projectId = 'PROJECT_ID',
  location = 'us-central1',
  model = 'gemini-1.5-flash-001',
  dataStoreId = 'DATASTORE_ID'
) {
  // Initialize Vertex with your Cloud project and location
  const vertexAI = new VertexAI({project: projectId, location: location});

  const generativeModelPreview = vertexAI.preview.getGenerativeModel({
    model: model,
    // The following parameters are optional
    // They can also be passed to individual content generation requests
    safetySettings: [
      {
        category: HarmCategory.HARM_CATEGORY_DANGEROUS_CONTENT,
        threshold: HarmBlockThreshold.BLOCK_MEDIUM_AND_ABOVE,
      },
    ],
    generationConfig: {maxOutputTokens: 256},
  });

  const vertexAIRetrievalTool = {
    retrieval: {
      vertexAiSearch: {
        datastore: `projects/${projectId}/locations/global/collections/default_collection/dataStores/${dataStoreId}`,
      },
      disableAttribution: false,
    },
  };

  const request = {
    contents: [{role: 'user', parts: [{text: 'Why is the sky blue?'}]}],
    tools: [vertexAIRetrievalTool],
  };

  const result = await generativeModelPreview.generateContent(request);
  const response = result.response;
  const groundingMetadata = response.candidates[0];
  console.log('Response: ', JSON.stringify(response.candidates[0]));
  console.log('GroundingMetadata is: ', JSON.stringify(groundingMetadata));
}

Java 

import com.google.cloud.vertexai.VertexAI;
import com.google.cloud.vertexai.api.GenerateContentResponse;
import com.google.cloud.vertexai.api.GroundingMetadata;
import com.google.cloud.vertexai.api.Retrieval;
import com.google.cloud.vertexai.api.Tool;
import com.google.cloud.vertexai.api.VertexAISearch;
import com.google.cloud.vertexai.generativeai.GenerativeModel;
import com.google.cloud.vertexai.generativeai.ResponseHandler;
import java.io.IOException;
import java.util.Collections;

public class GroundingWithPrivateData {
  public static void main(String[] args) throws IOException {
    // TODO(developer): Replace these variables before running the sample.
    String projectId = "your-google-cloud-project-id";
    String location = "us-central1";
    String modelName = "gemini-1.5-flash-001";
    String datastore = String.format(
        "projects/%s/locations/global/collections/default_collection/dataStores/%s",
        projectId, "datastore_id");

    groundWithPrivateData(projectId, location, modelName, datastore);
  }

  // A request whose response will be "grounded"
  // with information found in Vertex AI Search datastores.
  public static String groundWithPrivateData(String projectId, String location, String modelName,
                                             String datastoreId)
      throws IOException {
    // Initialize client that will be used to send requests.
    // This client only needs to be created once, and can be reused for multiple requests.
    try (VertexAI vertexAI = new VertexAI(projectId, location)) {
      Tool datastoreTool = Tool.newBuilder()
          .setRetrieval(
              Retrieval.newBuilder()
                  .setVertexAiSearch(VertexAISearch.newBuilder().setDatastore(datastoreId))
                  .setDisableAttribution(false))
          .build();

      GenerativeModel model = new GenerativeModel(modelName, vertexAI).withTools(
          Collections.singletonList(datastoreTool)
      );

      GenerateContentResponse response = model.generateContent(
          "How do I make an appointment to renew my driver's license?");

      GroundingMetadata groundingMetadata = response.getCandidates(0).getGroundingMetadata();
      String answer = ResponseHandler.getText(response);

      System.out.println("Answer: " + answer);
      System.out.println("Grounding metadata: " + groundingMetadata);

      return answer;
    }
  }
}

Langkah berikutnya

Untuk dokumentasi mendetail, lihat hal berikut: