【Amazon Bedrock Intelligent Prompt Routing】使ってみた！

$ pip install boto3

$ aws configure

import json
import boto3

# Bedrockクライアントの初期化
bedrock_runtime = boto3.client(
    "bedrock-runtime",
    region_name="us-east-1"
)

# プロンプトルーターのARNを指定
ROUTER_ARN = "<YOUR_PROMPT_ROUTER_ARN>"

# メッセージの作成
messages = [{
    "role": "user",
    "content": [
        {"text": "<YOUR_QUESTION>"}
    ]
}]

# プロンプトルーターを使用してモデルを呼び出し
response = bedrock_runtime.converse(
    modelId=ROUTER_ARN,
    messages=messages
)

# レスポンスの表示
print("Response:")
print(response["output"]["message"]["content"][0]["text"])

# 使用されたモデルの確認
print("\nUsed Model:")
print(json.dumps(response["trace"], indent=2))

$ python3 prompt-router-demo.py

Response:
Hello! How can I assist you today?

Used Model:
{
  "promptRouter": {
      "invokedModelId": "arn:aws:bedrock:us-east-1:414589897431:inference-profile/us.anthropic.claude-3-haiku-20240307-v1:0"
  }
}

Response:
Certainly! Here's a Python script that creates a DynamoDB table using the AWS SDK for Python (Boto3):

import boto3

def create_dynamodb_table(table_name):
    # Create a boto3 client for DynamoDB
    dynamodb = boto3.client('dynamodb')

    try:
        # Create the DynamoDB table
        response = dynamodb.create_table(
            TableName=table_name,
            KeySchema=[
                {
                    'AttributeName': 'id',
                    'KeyType': 'HASH'  # Partition key
                },
                {
                    'AttributeName': 'timestamp',
                    'KeyType': 'RANGE'  # Sort key
                }
            ],
            AttributeDefinitions=[
                {
                    'AttributeName': 'id',
                    'AttributeType': 'S'  # String
                },
                {
                    'AttributeName': 'timestamp',
                    'AttributeType': 'N'  # Number
                }
            ],
            ProvisionedThroughput={
                'ReadCapacityUnits': 5,
                'WriteCapacityUnits': 5
            }
        )

        print(f"Table {table_name} created successfully.")
        return response
    except dynamodb.exceptions.ResourceInUseException:
        print(f"Table {table_name} already exists.")
    except Exception as e:
        print(f"Error creating table: {str(e)}")

# Example usage
table_name = 'MyDynamoDBTable'
create_dynamodb_table(table_name)

This script does the following:

1. We import the boto3 library, which is the AWS SDK for Python.
2. We define a function create_dynamodb_table that takes a table_name as an argument.
3. Inside the function, we create a boto3 client for DynamoDB.
4. We use the create_table method of the DynamoDB client to create the table. We specify:
 - The table name
 - The key schema (in this example, we're using a composite key with 'id' as the partition key and 'timestamp' as the sort key)
 - The attribute definitions for the key attributes
 - The provisioned throughput for read and write capacity units

5. If the table is created successfully, we print a success message and return the response.
6. We include error handling for cases where the table already exists or if there's another error during creation.
7. At the end, we provide an example usage of the function.

To use this script:
1. Make sure you have the AWS CLI configured with your credentials or have the appropriate IAM role attached to your EC2 instance if you're running this on AWS.
2. Install boto3 if you haven't already: pip install boto3
3. Run the script. It will create a DynamoDB table with the name 'MyDynamoDBTable' (or whatever name you specify).

Remember to adjust the table schema, attribute definitions, and provisioned throughput according to your specific requirements.

Used Model:
{
  "promptRouter": {
  "invokedModelId": "arn:aws:bedrock:us-east-1:414589897431:inference-profile/us.anthropic.claude-3-5-sonnet-20240620-v1:0"
  }
}

$ aws bedrock list-prompt-routers

{
    "promptRouterSummaries": [
        {
            "promptRouterName": "Anthropic Prompt Router",
            "routingCriteria": {
                "responseQualityDifference": 0.26
            },
            "description": "Routes requests among models in the Claude family",
            "createdAt": "2024-11-20T00:00:00+00:00",
            "updatedAt": "2024-11-20T00:00:00+00:00",
            "promptRouterArn": "arn:aws:bedrock:us-east-1:123412341234:default-prompt-router/anthropic.claude:1",
            "models": [
                {
                    "modelArn": "arn:aws:bedrock:us-east-1:123412341234:inference-profile/us.anthropic.claude-3-haiku-20240307-v1:0"
                },
                {
                    "modelArn": "arn:aws:bedrock:us-east-1:123412341234:inference-profile/us.anthropic.claude-3-5-sonnet-20240620-v1:0"
                }
            ],
            "fallbackModel": {
                "modelArn": "arn:aws:bedrock:us-east-1:123412341234:inference-profile/us.anthropic.claude-3-5-sonnet-20240620-v1:0"
            },
            "status": "AVAILABLE",
            "type": "default"
        },
        {
            "promptRouterName": "Meta Prompt Router",
            "routingCriteria": {
                "responseQualityDifference": 0.0
            },
            "description": "Routes requests among models in the LLaMA family",
            "createdAt": "2024-11-20T00:00:00+00:00",
            "updatedAt": "2024-11-20T00:00:00+00:00",
            "promptRouterArn": "arn:aws:bedrock:us-east-1:123412341234:default-prompt-router/meta.llama:1",
            "models": [
                {
                    "modelArn": "arn:aws:bedrock:us-east-1:123412341234:inference-profile/us.meta.llama3-1-8b-instruct-v1:0"
                },
                {
                    "modelArn": "arn:aws:bedrock:us-east-1:123412341234:inference-profile/us.meta.llama3-1-70b-instruct-v1:0"
                }
            ],
            "fallbackModel": {
                "modelArn": "arn:aws:bedrock:us-east-1:123412341234:inference-profile/us.meta.llama3-1-70b-instruct-v1:0"
            },
            "status": "AVAILABLE",
            "type": "default"
        }
    ]
}

$ aws bedrock get-prompt-router --prompt-router-arn arn:aws:bedrock:us-east-1:123412341234:default-prompt-router/meta.llama:1

{
    "promptRouterName": "Meta Prompt Router",
    "models": [
        {
            "modelArn": "arn:aws:bedrock:us-east-1:123412341234:inference-profile/us.meta.llama3-1-8b-instruct-v1:0"
        },
        {
            "modelArn": "arn:aws:bedrock:us-east-1:123412341234:inference-profile/us.meta.llama3-1-70b-instruct-v1:0"
        }
    ],
    "fallbackModel": {
        "modelArn": "arn:aws:bedrock:us-east-1:123412341234:inference-profile/us.meta.llama3-1-70b-instruct-v1:0"
    },
    "status": "AVAILABLE"
}