SDK for JavaScript (v3) を使用した Amazon Bedrock ランタイムの例 - AWS SDK コード例

Doc AWS SDK Examples リポジトリには、他にも SDK の例があります。 AWS GitHub

翻訳は機械翻訳により提供されています。提供された翻訳内容と英語版の間で齟齬、不一致または矛盾がある場合、英語版が優先します。

SDK for JavaScript (v3) を使用した Amazon Bedrock ランタイムの例

次のコード例は、Amazon Bedrock ランタイムで AWS SDK for JavaScript (v3) を使用してアクションを実行し、一般的なシナリオを実装する方法を示しています。

アクションはより大きなプログラムからのコードの抜粋であり、コンテキスト内で実行する必要があります。アクションは個々のサービス機能を呼び出す方法を示していますが、関連するシナリオやサービス間の例ではアクションのコンテキストが確認できます。

「シナリオ」は、同じサービス内で複数の関数を呼び出して、特定のタスクを実行する方法を示すコード例です。

各例には、 へのリンクが含まれています。このリンクには GitHub、コンテキスト内でコードをセットアップして実行する方法の手順が記載されています。

開始方法

次のコード例は、Amazon Bedrock の使用を開始する方法を示しています。

SDK for JavaScript (v3)
注記

の詳細については、「」を参照してください GitHub。AWS コード例リポジトリ で全く同じ例を見つけて、設定と実行の方法を確認してください。

// Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. // SPDX-License-Identifier: Apache-2.0 /** * @typedef {Object} Content * @property {string} text * * @typedef {Object} Usage * @property {number} input_tokens * @property {number} oputput_tokens * * @typedef {Object} ResponseBody * @property {Content[]} content * @property {Usage} usage */ import { fileURLToPath } from "url"; import { BedrockRuntimeClient, InvokeModelCommand, } from "@aws-sdk/client-bedrock-runtime"; const AWS_REGION = "us-east-1"; const MODEL_ID = "anthropic.claude-3-haiku-20240307-v1:0"; const PROMPT = "Hi. In a short paragraph, explain what you can do."; const hello = async () => { console.log("=".repeat(35)); console.log("Welcome to the Amazon Bedrock demo!"); console.log("=".repeat(35)); console.log("Model: Anthropic Claude 3 Haiku"); console.log(`Prompt: ${PROMPT}\n`); console.log("Invoking model...\n"); // Create a new Bedrock Runtime client instance. const client = new BedrockRuntimeClient({ region: AWS_REGION }); // Prepare the payload for the model. const payload = { anthropic_version: "bedrock-2023-05-31", max_tokens: 1000, messages: [{ role: "user", content: [{ type: "text", text: PROMPT }] }], }; // Invoke Claude with the payload and wait for the response. const apiResponse = await client.send( new InvokeModelCommand({ contentType: "application/json", body: JSON.stringify(payload), modelId: MODEL_ID, }), ); // Decode and return the response(s) const decodedResponseBody = new TextDecoder().decode(apiResponse.body); /** @type {ResponseBody} */ const responseBody = JSON.parse(decodedResponseBody); const responses = responseBody.content; if (responses.length === 1) { console.log(`Response: ${responses[0].text}`); } else { console.log("Haiku returned multiple responses:"); console.log(responses); } console.log(`\nNumber of input tokens: ${responseBody.usage.input_tokens}`); console.log(`Number of output tokens: ${responseBody.usage.output_tokens}`); }; if (process.argv[1] === fileURLToPath(import.meta.url)) { await hello(); }
  • API の詳細については、「 API リファレンスInvokeModel」の「」を参照してください。 AWS SDK for JavaScript

モデルを呼び出す例

次のコード例は、テキスト生成のために Amazon Bedrock で AI21 Labs Jurassic-2 モデルを呼び出す方法を示しています。

SDK for JavaScript (v3)
注記

の詳細については、「」を参照してください GitHub。AWS コード例リポジトリ で全く同じ例を見つけて、設定と実行の方法を確認してください。

AI21 Labs Jurassic-2 基盤モデルを呼び出して、テキストを生成します。

// Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. // SPDX-License-Identifier: Apache-2.0 import { fileURLToPath } from "url"; import { FoundationModels } from "../../config/foundation_models.js"; import { BedrockRuntimeClient, InvokeModelCommand, } from "@aws-sdk/client-bedrock-runtime"; /** * @typedef {Object} Data * @property {string} text * * @typedef {Object} Completion * @property {Data} data * * @typedef {Object} ResponseBody * @property {Completion[]} completions */ /** * Invokes an AI21 Labs Jurassic-2 model. * * @param {string} prompt - The input text prompt for the model to complete. * @param {string} [modelId] - The ID of the model to use. Defaults to "ai21.j2-mid-v1". */ export const invokeModel = async (prompt, modelId = "ai21.j2-mid-v1") => { // Create a new Bedrock Runtime client instance. const client = new BedrockRuntimeClient({ region: "us-east-1" }); // Prepare the payload for the model. const payload = { prompt, maxTokens: 500, temperature: 0.5, }; // Invoke the model with the payload and wait for the response. const command = new InvokeModelCommand({ contentType: "application/json", body: JSON.stringify(payload), modelId, }); const apiResponse = await client.send(command); // Decode and return the response(s). const decodedResponseBody = new TextDecoder().decode(apiResponse.body); /** @type {ResponseBody} */ const responseBody = JSON.parse(decodedResponseBody); return responseBody.completions[0].data.text; }; // Invoke the function if this file was run directly. if (process.argv[1] === fileURLToPath(import.meta.url)) { const prompt = 'Complete the following in one sentence: "Once upon a time..."'; const modelId = FoundationModels.JURASSIC2_MID.modelId; console.log(`Prompt: ${prompt}`); console.log(`Model ID: ${modelId}`); try { console.log("-".repeat(53)); const response = await invokeModel(prompt, modelId); console.log(response); } catch (err) { console.log(err); } }
  • API の詳細については、「 API リファレンスInvokeModel」の「」を参照してください。 AWS SDK for JavaScript

次のコード例は、Amazon Bedrock で Amazon Titan Text G1 を呼び出してテキストを生成する方法を示しています。

SDK for JavaScript (v3)
注記

の詳細については、「」を参照してください GitHub。AWS コード例リポジトリ で全く同じ例を見つけて、設定と実行の方法を確認してください。

Amazon Titan Text G1 基盤モデルを呼び出してテキストを生成します。

// Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. // SPDX-License-Identifier: Apache-2.0 import { fileURLToPath } from "url"; import { FoundationModels } from "../../config/foundation_models.js"; import { BedrockRuntimeClient, InvokeModelCommand, } from "@aws-sdk/client-bedrock-runtime"; /** * @typedef {Object} ResponseBody * @property {Object[]} results */ /** * Invokes an Amazon Titan Text generation model. * * @param {string} prompt - The input text prompt for the model to complete. * @param {string} [modelId] - The ID of the model to use. Defaults to "amazon.titan-text-express-v1". */ export const invokeModel = async ( prompt, modelId = "amazon.titan-text-express-v1", ) => { // Create a new Bedrock Runtime client instance. const client = new BedrockRuntimeClient({ region: "us-east-1" }); // Prepare the payload for the model. const payload = { inputText: prompt, textGenerationConfig: { maxTokenCount: 4096, stopSequences: [], temperature: 0, topP: 1, }, }; // Invoke the model with the payload and wait for the response. const command = new InvokeModelCommand({ contentType: "application/json", body: JSON.stringify(payload), modelId, }); const apiResponse = await client.send(command); // Decode and return the response. const decodedResponseBody = new TextDecoder().decode(apiResponse.body); /** @type {ResponseBody} */ const responseBody = JSON.parse(decodedResponseBody); return responseBody.results[0].outputText; }; // Invoke the function if this file was run directly. if (process.argv[1] === fileURLToPath(import.meta.url)) { const prompt = 'Complete the following in one sentence: "Once upon a time..."'; const modelId = FoundationModels.TITAN_TEXT_G1_EXPRESS.modelId; console.log(`Prompt: ${prompt}`); console.log(`Model ID: ${modelId}`); try { console.log("-".repeat(53)); const response = await invokeModel(prompt, modelId); console.log(response); } catch (err) { console.log(err); } }
  • API の詳細については、「 API リファレンスInvokeModel」の「」を参照してください。 AWS SDK for JavaScript

次のコード例は、Amazon Bedrock で Anthropic Claude 2.x を呼び出し、レスポンスストリームをリアルタイムで処理する方法を示しています。

SDK for JavaScript (v3)
注記

の詳細については、「」を参照してください GitHub。AWS コード例リポジトリ で全く同じ例を見つけて、設定と実行の方法を確認してください。

Anthropic Claude を呼び出し、レスポンスストリームを処理します。

// Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. // SPDX-License-Identifier: Apache-2.0 import { fileURLToPath } from "url"; import { FoundationModels } from "../../config/foundation_models.js"; import { BedrockRuntimeClient, InvokeModelCommand, InvokeModelWithResponseStreamCommand, } from "@aws-sdk/client-bedrock-runtime"; /** * @typedef {Object} ResponseContent * @property {string} text * * @typedef {Object} MessagesResponseBody * @property {ResponseContent[]} content * * @typedef {Object} Delta * @property {string} text * * @typedef {Object} Message * @property {string} role * * @typedef {Object} Chunk * @property {string} type * @property {Delta} delta * @property {Message} message */ /** * Invokes Anthropic Claude 3 using the Messages API. * * To learn more about the Anthropic Messages API, go to: * https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-anthropic-claude-messages.html * * @param {string} prompt - The input text prompt for the model to complete. * @param {string} [modelId] - The ID of the model to use. Defaults to "anthropic.claude-3-haiku-20240307-v1:0". */ export const invokeModel = async ( prompt, modelId = "anthropic.claude-3-haiku-20240307-v1:0", ) => { // Create a new Bedrock Runtime client instance. const client = new BedrockRuntimeClient({ region: "us-east-1" }); // Prepare the payload for the model. const payload = { anthropic_version: "bedrock-2023-05-31", max_tokens: 1000, messages: [ { role: "user", content: [{ type: "text", text: prompt }], }, ], }; // Invoke Claude with the payload and wait for the response. const command = new InvokeModelCommand({ contentType: "application/json", body: JSON.stringify(payload), modelId, }); const apiResponse = await client.send(command); // Decode and return the response(s) const decodedResponseBody = new TextDecoder().decode(apiResponse.body); /** @type {MessagesResponseBody} */ const responseBody = JSON.parse(decodedResponseBody); return responseBody.content[0].text; }; /** * Invokes Anthropic Claude 3 and processes the response stream. * * To learn more about the Anthropic Messages API, go to: * https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-anthropic-claude-messages.html * * @param {string} prompt - The input text prompt for the model to complete. * @param {string} [modelId] - The ID of the model to use. Defaults to "anthropic.claude-3-haiku-20240307-v1:0". */ export const invokeModelWithResponseStream = async ( prompt, modelId = "anthropic.claude-3-haiku-20240307-v1:0", ) => { // Create a new Bedrock Runtime client instance. const client = new BedrockRuntimeClient({ region: "us-east-1" }); // Prepare the payload for the model. const payload = { anthropic_version: "bedrock-2023-05-31", max_tokens: 1000, messages: [ { role: "user", content: [{ type: "text", text: prompt }], }, ], }; // Invoke Claude with the payload and wait for the API to respond. const command = new InvokeModelWithResponseStreamCommand({ contentType: "application/json", body: JSON.stringify(payload), modelId, }); const apiResponse = await client.send(command); let completeMessage = ""; // Decode and process the response stream for await (const item of apiResponse.body) { /** @type Chunk */ const chunk = JSON.parse(new TextDecoder().decode(item.chunk.bytes)); const chunk_type = chunk.type; if (chunk_type === "content_block_delta") { const text = chunk.delta.text; completeMessage = completeMessage + text; process.stdout.write(text); } } // Return the final response return completeMessage; }; // Invoke the function if this file was run directly. if (process.argv[1] === fileURLToPath(import.meta.url)) { const prompt = 'Write a paragraph starting with: "Once upon a time..."'; const modelId = FoundationModels.CLAUDE_3_HAIKU.modelId; console.log(`Prompt: ${prompt}`); console.log(`Model ID: ${modelId}`); try { console.log("-".repeat(53)); const response = await invokeModel(prompt, modelId); console.log("\n" + "-".repeat(53)); console.log("Final structured response:"); console.log(response); } catch (err) { console.log(`\n${err}`); } }
  • API の詳細については、「 API リファレンスInvokeModelWithResponseStream」の「」を参照してください。 AWS SDK for JavaScript

次のコード例は、Amazon Bedrock で Anthropic Claude 2.x を呼び出してテキストを生成する方法を示しています。

SDK for JavaScript (v3)
注記

の詳細については、「」を参照してください GitHub。AWS コード例リポジトリ で全く同じ例を見つけて、設定と実行の方法を確認してください。

Anthropic Claude 2 基盤モデルを呼び出して、テキストを生成します。

// Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. // SPDX-License-Identifier: Apache-2.0 import { fileURLToPath } from "url"; import { FoundationModels } from "../../config/foundation_models.js"; import { BedrockRuntimeClient, InvokeModelCommand, } from "@aws-sdk/client-bedrock-runtime"; /** * @typedef {Object} ResponseContent * @property {string} text */ /** * @typedef {Object} MessagesResponseBody * @property {ResponseContent[]} content */ /** * @typedef {Object} TextCompletionsResponseBody * @property {completion} text */ /** * Invokes Anthropic Claude 2.x using the Messages API. * * To learn more about the Anthropic Messages API, go to: * https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-anthropic-claude-messages.html * * @param {string} prompt - The input text prompt for the model to complete. * @param {string} [modelId] - The ID of the model to use. Defaults to "anthropic.claude-v2". */ export const invokeModel = async (prompt, modelId = "anthropic.claude-v2") => { // Create a new Bedrock Runtime client instance. const client = new BedrockRuntimeClient({ region: "us-east-1" }); // Prepare the payload for the Messages API request. const payload = { anthropic_version: "bedrock-2023-05-31", max_tokens: 1000, messages: [ { role: "user", content: [{ type: "text", text: prompt }], }, ], }; // Invoke Claude with the payload and wait for the response. const command = new InvokeModelCommand({ contentType: "application/json", body: JSON.stringify(payload), modelId, }); const apiResponse = await client.send(command); // Decode and return the response(s) const decodedResponseBody = new TextDecoder().decode(apiResponse.body); /** @type {MessagesResponseBody} */ const responseBody = JSON.parse(decodedResponseBody); return responseBody.content[0].text; }; /** * Invokes Anthropic Claude 2.x using the Text Completions API. * * To learn more about the Anthropic Text Completions API, go to: * https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-anthropic-claude-text-completion.html * * @param {string} prompt - The input text prompt for the model to complete. * @param {string} [modelId] - The ID of the model to use. Defaults to "anthropic.claude-v2". */ export const invokeTextCompletionsApi = async ( prompt, modelId = "anthropic.claude-v2", ) => { // Create a new Bedrock Runtime client instance. const client = new BedrockRuntimeClient({ region: "us-east-1" }); // Prepare the payload for the Text Completions API, using the required prompt template. const enclosedPrompt = `Human: ${prompt}\n\nAssistant:`; const payload = { prompt: enclosedPrompt, max_tokens_to_sample: 500, temperature: 0.5, stop_sequences: ["\n\nHuman:"], }; // Invoke Claude with the payload and wait for the response. const command = new InvokeModelCommand({ contentType: "application/json", body: JSON.stringify(payload), modelId, }); const apiResponse = await client.send(command); // Decode and return the response. const decoded = new TextDecoder().decode(apiResponse.body); /** @type {TextCompletionsResponseBody} */ const responseBody = JSON.parse(decoded); return responseBody.completion; }; // Invoke the function if this file was run directly. if (process.argv[1] === fileURLToPath(import.meta.url)) { const prompt = 'Complete the following in one sentence: "Once upon a time..."'; const modelId = FoundationModels.CLAUDE_2.modelId; console.log(`Prompt: ${prompt}`); console.log(`Model ID: ${modelId}`); try { console.log("-".repeat(53)); console.log("Using the Messages API:"); const response = await invokeModel(prompt, modelId); console.log(response); } catch (err) { console.log(err); } try { console.log("-".repeat(53)); console.log("Using the Text Completions API:"); const response = await invokeTextCompletionsApi(prompt, modelId); console.log(response); } catch (err) { console.log(err); } }
  • API の詳細については、「 API リファレンスInvokeModel」の「」を参照してください。 AWS SDK for JavaScript

次のコード例は、Amazon Bedrock で Anthropic Claude 3 を呼び出してテキストを生成する方法を示しています。

SDK for JavaScript (v3)
注記

の詳細については、「」を参照してください GitHub。AWS コード例リポジトリ で全く同じ例を見つけて、設定と実行の方法を確認してください。

Anthropic Claude 3 を呼び出してテキストを生成します。

// Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. // SPDX-License-Identifier: Apache-2.0 import { fileURLToPath } from "url"; import { FoundationModels } from "../../config/foundation_models.js"; import { BedrockRuntimeClient, InvokeModelCommand, InvokeModelWithResponseStreamCommand, } from "@aws-sdk/client-bedrock-runtime"; /** * @typedef {Object} ResponseContent * @property {string} text * * @typedef {Object} MessagesResponseBody * @property {ResponseContent[]} content * * @typedef {Object} Delta * @property {string} text * * @typedef {Object} Message * @property {string} role * * @typedef {Object} Chunk * @property {string} type * @property {Delta} delta * @property {Message} message */ /** * Invokes Anthropic Claude 3 using the Messages API. * * To learn more about the Anthropic Messages API, go to: * https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-anthropic-claude-messages.html * * @param {string} prompt - The input text prompt for the model to complete. * @param {string} [modelId] - The ID of the model to use. Defaults to "anthropic.claude-3-haiku-20240307-v1:0". */ export const invokeModel = async ( prompt, modelId = "anthropic.claude-3-haiku-20240307-v1:0", ) => { // Create a new Bedrock Runtime client instance. const client = new BedrockRuntimeClient({ region: "us-east-1" }); // Prepare the payload for the model. const payload = { anthropic_version: "bedrock-2023-05-31", max_tokens: 1000, messages: [ { role: "user", content: [{ type: "text", text: prompt }], }, ], }; // Invoke Claude with the payload and wait for the response. const command = new InvokeModelCommand({ contentType: "application/json", body: JSON.stringify(payload), modelId, }); const apiResponse = await client.send(command); // Decode and return the response(s) const decodedResponseBody = new TextDecoder().decode(apiResponse.body); /** @type {MessagesResponseBody} */ const responseBody = JSON.parse(decodedResponseBody); return responseBody.content[0].text; }; /** * Invokes Anthropic Claude 3 and processes the response stream. * * To learn more about the Anthropic Messages API, go to: * https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-anthropic-claude-messages.html * * @param {string} prompt - The input text prompt for the model to complete. * @param {string} [modelId] - The ID of the model to use. Defaults to "anthropic.claude-3-haiku-20240307-v1:0". */ export const invokeModelWithResponseStream = async ( prompt, modelId = "anthropic.claude-3-haiku-20240307-v1:0", ) => { // Create a new Bedrock Runtime client instance. const client = new BedrockRuntimeClient({ region: "us-east-1" }); // Prepare the payload for the model. const payload = { anthropic_version: "bedrock-2023-05-31", max_tokens: 1000, messages: [ { role: "user", content: [{ type: "text", text: prompt }], }, ], }; // Invoke Claude with the payload and wait for the API to respond. const command = new InvokeModelWithResponseStreamCommand({ contentType: "application/json", body: JSON.stringify(payload), modelId, }); const apiResponse = await client.send(command); let completeMessage = ""; // Decode and process the response stream for await (const item of apiResponse.body) { /** @type Chunk */ const chunk = JSON.parse(new TextDecoder().decode(item.chunk.bytes)); const chunk_type = chunk.type; if (chunk_type === "content_block_delta") { const text = chunk.delta.text; completeMessage = completeMessage + text; process.stdout.write(text); } } // Return the final response return completeMessage; }; // Invoke the function if this file was run directly. if (process.argv[1] === fileURLToPath(import.meta.url)) { const prompt = 'Write a paragraph starting with: "Once upon a time..."'; const modelId = FoundationModels.CLAUDE_3_HAIKU.modelId; console.log(`Prompt: ${prompt}`); console.log(`Model ID: ${modelId}`); try { console.log("-".repeat(53)); const response = await invokeModel(prompt, modelId); console.log("\n" + "-".repeat(53)); console.log("Final structured response:"); console.log(response); } catch (err) { console.log(`\n${err}`); } }
  • API の詳細については、「 API リファレンスInvokeModel」の「」を参照してください。 AWS SDK for JavaScript

次のコード例は、Amazon Bedrock で Anthropic Claude Instant モデルを呼び出してテキストを生成する方法を示しています。

SDK for JavaScript (v3)
注記

の詳細については、「」を参照してください GitHub。AWS コード例リポジトリ で全く同じ例を見つけて、設定と実行の方法を確認してください。

Anthropic Claude Instant 基盤モデルを呼び出してテキストを生成します。

// Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. // SPDX-License-Identifier: Apache-2.0 import { fileURLToPath } from "url"; import { FoundationModels } from "../../config/foundation_models.js"; import { BedrockRuntimeClient, InvokeModelCommand, } from "@aws-sdk/client-bedrock-runtime"; /** * @typedef {Object} Content * @property {string} text * * @typedef {Object} MessageApiResponse * @property {Content[]} content */ /** * @typedef {Object} TextCompletionApiResponse * @property {string} completion */ /** * Invokes Anthropic Claude Instant using the Messages API. * * To learn more about the Anthropic Messages API, go to: * https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-anthropic-claude-messages.html * * @param {string} prompt - The input text prompt for the model to complete. * @param {string} [modelId] - The ID of the model to use. Defaults to "anthropic.claude-instant-v1". */ export const invokeModel = async ( prompt, modelId = "anthropic.claude-instant-v1", ) => { // Create a new Bedrock Runtime client instance. const client = new BedrockRuntimeClient({ region: "us-east-1" }); // Prepare the payload for the Messages API request. const payload = { anthropic_version: "bedrock-2023-05-31", max_tokens: 1000, messages: [ { role: "user", content: [{ type: "text", text: prompt }], }, ], }; // Invoke Claude with the payload and wait for the response. const command = new InvokeModelCommand({ contentType: "application/json", body: JSON.stringify(payload), modelId, }); const apiResponse = await client.send(command); // Decode and return the response(s) const decodedResponseBody = new TextDecoder().decode(apiResponse.body); /** @type {MessageApiResponse} */ const responseBody = JSON.parse(decodedResponseBody); return responseBody.content[0].text; }; /** * Invokes Anthropic Claude Instant using the Text Completions API. * * To learn more about the Anthropic Text Completions API, go to: * https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-anthropic-claude-text-completion.html * * @param {string} prompt - The input text prompt for the model to complete. * @param {string} [modelId] - The ID of the model to use. Defaults to "anthropic.claude-instant-v1". */ export const invokeTextCompletionsApi = async ( prompt, modelId = "anthropic.claude-instant-v1", ) => { // Create a new Bedrock Runtime client instance. const client = new BedrockRuntimeClient({ region: "us-east-1" }); // Prepare the payload for the Text Completions API, using the required prompt template. const enclosedPrompt = `Human: ${prompt}\n\nAssistant:`; const payload = { prompt: enclosedPrompt, max_tokens_to_sample: 500, temperature: 0.5, stop_sequences: ["\n\nHuman:"], }; // Invoke Claude with the payload and wait for the response. const command = new InvokeModelCommand({ contentType: "application/json", body: JSON.stringify(payload), modelId, }); const apiResponse = await client.send(command); // Decode and return the response. const decoded = new TextDecoder().decode(apiResponse.body); /** @type {TextCompletionApiResponse} */ const responseBody = JSON.parse(decoded); return responseBody.completion; }; // Invoke the function if this file was run directly. if (process.argv[1] === fileURLToPath(import.meta.url)) { const prompt = 'Complete the following in one sentence: "Once upon a time..."'; const modelId = FoundationModels.CLAUDE_INSTANT.modelId; console.log(`Prompt: ${prompt}`); console.log(`Model ID: ${modelId}`); try { console.log("-".repeat(53)); console.log("Using the Messages API:"); const response = await invokeModel(prompt, modelId); console.log(response); } catch (err) { console.log(err); } try { console.log("-".repeat(53)); console.log("Using the Text Completions API:"); const response = await invokeTextCompletionsApi(prompt, modelId); console.log(response); } catch (err) { console.log(err); } }
  • API の詳細については、「 API リファレンスInvokeModel」の「」を参照してください。 AWS SDK for JavaScript

次のコード例は、Meta Llama 2 へのプロンプトの送信とレスポンスの出力を開始する方法を示しています。

SDK for JavaScript (v3)
注記

の詳細については、「」を参照してください GitHub。AWS コード例リポジトリ で全く同じ例を見つけて、設定と実行の方法を確認してください。

最初のプロンプトを Meta Llama 2 に送信します。

// Send a prompt to Meta Llama 2 and print the response. import { BedrockRuntimeClient, InvokeModelCommand, } from "@aws-sdk/client-bedrock-runtime"; // Create a Bedrock Runtime client in the AWS Region of your choice. const client = new BedrockRuntimeClient({ region: "us-west-2" }); // Set the model ID, e.g., Llama 2 Chat 13B. const modelId = "meta.llama2-13b-chat-v1"; // Define the user message to send. const userMessage = "Describe the purpose of a 'hello world' program in one sentence."; // Embed the message in Llama 2's prompt format. const prompt = `<s>[INST] ${userMessage} [/INST]`; // Format the request payload using the model's native structure. const request = { prompt, // Optional inference parameters: max_gen_len: 512, temperature: 0.5, top_p: 0.9, }; // Encode and send the request. const response = await client.send( new InvokeModelCommand({ contentType: "application/json", body: JSON.stringify(request), modelId, }), ); // Decode the native response body. /** @type {{ generation: string }} */ const nativeResponse = JSON.parse(new TextDecoder().decode(response.body)); // Extract and print the generated text. const responseText = nativeResponse.generation; console.log(responseText); // Learn more about the Llama 2 prompt format at: // https://llama.meta.com/docs/model-cards-and-prompt-formats/meta-llama-2
  • API の詳細については、「 API リファレンスInvokeModel」の「」を参照してください。 AWS SDK for JavaScript

次のコード例は、Meta Llama 2 にプロンプトを送信し、レスポンスストリームをリアルタイムで出力する方法を示しています。

SDK for JavaScript (v3)
注記

の詳細については、「」を参照してください GitHub。AWS コード例リポジトリ で全く同じ例を見つけて、設定と実行の方法を確認してください。

最初のプロンプトを Meta Llama 3 に送信します。

// Send a prompt to Meta Llama 2 and print the response stream in real-time. import { BedrockRuntimeClient, InvokeModelWithResponseStreamCommand, } from "@aws-sdk/client-bedrock-runtime"; // Create a Bedrock Runtime client in the AWS Region of your choice. const client = new BedrockRuntimeClient({ region: "us-west-2" }); // Set the model ID, e.g., Llama 2 Chat 13B. const modelId = "meta.llama2-13b-chat-v1"; // Define the user message to send. const userMessage = "Describe the purpose of a 'hello world' program in one sentence."; // Embed the message in Llama 2's prompt format. const prompt = `<s>[INST] ${userMessage} [/INST]`; // Format the request payload using the model's native structure. const request = { prompt, // Optional inference parameters: max_gen_len: 512, temperature: 0.5, top_p: 0.9, }; // Encode and send the request. const responseStream = await client.send( new InvokeModelWithResponseStreamCommand({ contentType: "application/json", body: JSON.stringify(request), modelId, }), ); // Extract and print the response stream in real-time. for await (const event of responseStream.body) { /** @type {{ generation: string }} */ const chunk = JSON.parse(new TextDecoder().decode(event.chunk.bytes)); if (chunk.generation) { process.stdout.write(chunk.generation); } } // Learn more about the Llama 3 prompt format at: // https://llama.meta.com/docs/model-cards-and-prompt-formats/meta-llama-3/#special-tokens-used-with-meta-llama-3
  • API の詳細については、「 API リファレンスInvokeModelWithResponseStream」の「」を参照してください。 AWS SDK for JavaScript

次のコード例は、Meta Llama 3 へのプロンプトの送信とレスポンスの出力を開始する方法を示しています。

SDK for JavaScript (v3)
注記

の詳細については、「」を参照してください GitHub。AWS コード例リポジトリ で全く同じ例を見つけて、設定と実行の方法を確認してください。

最初のプロンプトを Meta Llama 3 に送信します。

// Send a prompt to Meta Llama 3 and print the response. import { BedrockRuntimeClient, InvokeModelCommand, } from "@aws-sdk/client-bedrock-runtime"; // Create a Bedrock Runtime client in the AWS Region of your choice. const client = new BedrockRuntimeClient({ region: "us-west-2" }); // Set the model ID, e.g., Llama 3 8B Instruct. const modelId = "meta.llama3-8b-instruct-v1:0"; // Define the user message to send. const userMessage = "Describe the purpose of a 'hello world' program in one sentence."; // Embed the message in Llama 3's prompt format. const prompt = ` <|begin_of_text|> <|start_header_id|>user<|end_header_id|> ${userMessage} <|eot_id|> <|start_header_id|>assistant<|end_header_id|> `; // Format the request payload using the model's native structure. const request = { prompt, // Optional inference parameters: max_gen_len: 512, temperature: 0.5, top_p: 0.9, }; // Encode and send the request. const response = await client.send( new InvokeModelCommand({ contentType: "application/json", body: JSON.stringify(request), modelId, }), ); // Decode the native response body. /** @type {{ generation: string }} */ const nativeResponse = JSON.parse(new TextDecoder().decode(response.body)); // Extract and print the generated text. const responseText = nativeResponse.generation; console.log(responseText); // Learn more about the Llama 3 prompt format at: // https://llama.meta.com/docs/model-cards-and-prompt-formats/meta-llama-3/#special-tokens-used-with-meta-llama-3
  • API の詳細については、「 API リファレンスInvokeModel」の「」を参照してください。 AWS SDK for JavaScript

次のコード例は、Meta Llama 3 にプロンプトを送信し、レスポンスストリームをリアルタイムで出力する方法を示しています。

SDK for JavaScript (v3)
注記

の詳細については、「」を参照してください GitHub。AWS コード例リポジトリ で全く同じ例を見つけて、設定と実行の方法を確認してください。

最初のプロンプトを Meta Llama 3 に送信します。

// Send a prompt to Meta Llama 3 and print the response stream in real-time. import { BedrockRuntimeClient, InvokeModelWithResponseStreamCommand, } from "@aws-sdk/client-bedrock-runtime"; // Create a Bedrock Runtime client in the AWS Region of your choice. const client = new BedrockRuntimeClient({ region: "us-west-2" }); // Set the model ID, e.g., Llama 3 8B Instruct. const modelId = "meta.llama3-8b-instruct-v1:0"; // Define the user message to send. const userMessage = "Describe the purpose of a 'hello world' program in one sentence."; // Embed the message in Llama 3's prompt format. const prompt = ` <|begin_of_text|> <|start_header_id|>user<|end_header_id|> ${userMessage} <|eot_id|> <|start_header_id|>assistant<|end_header_id|> `; // Format the request payload using the model's native structure. const request = { prompt, // Optional inference parameters: max_gen_len: 512, temperature: 0.5, top_p: 0.9, }; // Encode and send the request. const responseStream = await client.send( new InvokeModelWithResponseStreamCommand({ contentType: "application/json", body: JSON.stringify(request), modelId, }), ); // Extract and print the response stream in real-time. for await (const event of responseStream.body) { /** @type {{ generation: string }} */ const chunk = JSON.parse(new TextDecoder().decode(event.chunk.bytes)); if (chunk.generation) { process.stdout.write(chunk.generation); } } // Learn more about the Llama 3 prompt format at: // https://llama.meta.com/docs/model-cards-and-prompt-formats/meta-llama-3/#special-tokens-used-with-meta-llama-3
  • API の詳細については、「 API リファレンスInvokeModelWithResponseStream」の「」を参照してください。 AWS SDK for JavaScript

次のコード例は、テキスト生成のために Amazon Bedrock で Mistral 7B モデルを呼び出す方法を示しています。

SDK for JavaScript (v3)
注記

の詳細については、「」を参照してください GitHub。AWS コード例リポジトリ で全く同じ例を見つけて、設定と実行の方法を確認してください。

Mistral 7B 基盤モデルを呼び出してテキストを生成します。

// Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. // SPDX-License-Identifier: Apache-2.0 import { fileURLToPath } from "url"; import { FoundationModels } from "../../config/foundation_models.js"; import { BedrockRuntimeClient, InvokeModelCommand, } from "@aws-sdk/client-bedrock-runtime"; /** * @typedef {Object} Output * @property {string} text * * @typedef {Object} ResponseBody * @property {Output[]} outputs */ /** * Invokes a Mistral 7B Instruct model. * * @param {string} prompt - The input text prompt for the model to complete. * @param {string} [modelId] - The ID of the model to use. Defaults to "mistral.mistral-7b-instruct-v0:2". */ export const invokeModel = async ( prompt, modelId = "mistral.mistral-7b-instruct-v0:2", ) => { // Create a new Bedrock Runtime client instance. const client = new BedrockRuntimeClient({ region: "us-east-1" }); // Mistral instruct models provide optimal results when embedding // the prompt into the following template: const instruction = `<s>[INST] ${prompt} [/INST]`; // Prepare the payload. const payload = { prompt: instruction, max_tokens: 500, temperature: 0.5, }; // Invoke the model with the payload and wait for the response. const command = new InvokeModelCommand({ contentType: "application/json", body: JSON.stringify(payload), modelId, }); const apiResponse = await client.send(command); // Decode and return the response. const decodedResponseBody = new TextDecoder().decode(apiResponse.body); /** @type {ResponseBody} */ const responseBody = JSON.parse(decodedResponseBody); return responseBody.outputs[0].text; }; // Invoke the function if this file was run directly. if (process.argv[1] === fileURLToPath(import.meta.url)) { const prompt = 'Complete the following in one sentence: "Once upon a time..."'; const modelId = FoundationModels.MISTRAL_7B.modelId; console.log(`Prompt: ${prompt}`); console.log(`Model ID: ${modelId}`); try { console.log("-".repeat(53)); const response = await invokeModel(prompt, modelId); console.log(response); } catch (err) { console.log(err); } }
  • API の詳細については、「 API リファレンスInvokeModel」の「」を参照してください。 AWS SDK for JavaScript

次のコード例は、テキスト生成のために Amazon Bedrock で Mixtral 8x7B モデルを呼び出す方法を示しています。

SDK for JavaScript (v3)
注記

の詳細については、「」を参照してください GitHub。AWS コード例リポジトリ で全く同じ例を見つけて、設定と実行の方法を確認してください。

Mixtral 8x7B 基盤モデルを呼び出してテキストを生成します。

// Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. // SPDX-License-Identifier: Apache-2.0 import { fileURLToPath } from "url"; import { FoundationModels } from "../../config/foundation_models.js"; import { BedrockRuntimeClient, InvokeModelCommand, } from "@aws-sdk/client-bedrock-runtime"; /** * @typedef {Object} Output * @property {string} text * * @typedef {Object} ResponseBody * @property {Output[]} outputs */ /** * Invokes a Mistral 8x7B Instruct model. * * @param {string} prompt - The input text prompt for the model to complete. * @param {string} [modelId] - The ID of the model to use. Defaults to "mistral.mixtral-8x7b-instruct-v0:1". */ export const invokeModel = async ( prompt, modelId = "mistral.mixtral-8x7b-instruct-v0:1", ) => { // Create a new Bedrock Runtime client instance. const client = new BedrockRuntimeClient({ region: "us-east-1" }); // Mistral instruct models provide optimal results when embedding // the prompt into the following template: const instruction = `<s>[INST] ${prompt} [/INST]`; // Prepare the payload. const payload = { prompt: instruction, max_tokens: 500, temperature: 0.5, }; // Invoke the model with the payload and wait for the response. const command = new InvokeModelCommand({ contentType: "application/json", body: JSON.stringify(payload), modelId, }); const apiResponse = await client.send(command); // Decode and return the response. const decodedResponseBody = new TextDecoder().decode(apiResponse.body); /** @type {ResponseBody} */ const responseBody = JSON.parse(decodedResponseBody); return responseBody.outputs[0].text; }; // Invoke the function if this file was run directly. if (process.argv[1] === fileURLToPath(import.meta.url)) { const prompt = 'Complete the following in one sentence: "Once upon a time..."'; const modelId = FoundationModels.MIXTRAL_8X7B.modelId; console.log(`Prompt: ${prompt}`); console.log(`Model ID: ${modelId}`); try { console.log("-".repeat(53)); const response = await invokeModel(prompt, modelId); console.log(response); } catch (err) { console.log(err); } }
  • API の詳細については、「 API リファレンスInvokeModel」の「」を参照してください。 AWS SDK for JavaScript

シナリオ

次のコード例は、Amazon Bedrock でさまざまな大規模言語モデル (LLMsを準備してプロンプトを送信する方法を示しています。

SDK for JavaScript (v3)
注記

の詳細については、「」を参照してください GitHub。AWS コード例リポジトリ で全く同じ例を見つけて、設定と実行の方法を確認してください。

// Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. // SPDX-License-Identifier: Apache-2.0 import { fileURLToPath } from "url"; import { Scenario, ScenarioAction, ScenarioInput, ScenarioOutput, } from "@aws-doc-sdk-examples/lib/scenario/index.js"; import { FoundationModels } from "../config/foundation_models.js"; /** * @typedef {Object} ModelConfig * @property {Function} module * @property {Function} invoker * @property {string} modelId * @property {string} modelName */ const greeting = new ScenarioOutput( "greeting", "Welcome to the Amazon Bedrock Runtime client demo!", { header: true }, ); const selectModel = new ScenarioInput("model", "First, select a model:", { type: "select", choices: Object.values(FoundationModels).map((model) => ({ name: model.modelName, value: model, })), }); const enterPrompt = new ScenarioInput("prompt", "Now, enter your prompt:", { type: "input", }); const printDetails = new ScenarioOutput( "print details", /** * @param {{ model: ModelConfig, prompt: string }} c */ (c) => console.log(`Invoking ${c.model.modelName} with '${c.prompt}'...`), { slow: false }, ); const invokeModel = new ScenarioAction( "invoke model", /** * @param {{ model: ModelConfig, prompt: string, response: string }} c */ async (c) => { const modelModule = await c.model.module(); const invoker = c.model.invoker(modelModule); c.response = await invoker(c.prompt, c.model.modelId); }, ); const printResponse = new ScenarioOutput( "print response", /** * @param {{ response: string }} c */ (c) => c.response, { slow: false }, ); const scenario = new Scenario("Amazon Bedrock Runtime Demo", [ greeting, selectModel, enterPrompt, printDetails, invokeModel, printResponse, ]); if (process.argv[1] === fileURLToPath(import.meta.url)) { scenario.run(); }