具有知识的代理

代码

cookbook/models/aws/claude/knowledge.py

from agno.agent import Agent
from agno.knowledge.pdf_url import PDFUrlKnowledgeBase
from agno.models.aws import Claude
from agno.vectordb.pgvector import PgVector

db_url = "postgresql+psycopg://ai:ai@localhost:5532/ai"

knowledge_base = PDFUrlKnowledgeBase(
    urls=["https://agno-public.s3.amazonaws.com/recipes/ThaiRecipes.pdf"],
    vector_db=PgVector(table_name="recipes", db_url=db_url),
)
knowledge_base.load(recreate=True)  # 首次运行后注释掉此行

agent = Agent(
    model=Claude(id="claude-3-5-sonnet-20241022"),
    knowledge=knowledge_base,
    show_tool_calls=True,
)
agent.print_response("How to make Thai curry?", markdown=True)

用法

1. 创建虚拟环境

   打开 Terminal 并创建一个 python 虚拟环境。

   Mac

   python3 -m venv .venv
   source .venv/bin/activate
   

2. 设置您的 AWS 凭证

   export AWS_ACCESS_KEY_ID=***
   export AWS_SECRET_ACCESS_KEY=***
   export AWS_REGION=***
   

3. 安装库

   pip install -U anthropic[bedrock] sqlalchemy pgvector pypdf openai psycopg agno
   

4. 运行 PgVector

   docker run -d \
     -e POSTGRES_DB=ai \
     -e POSTGRES_USER=ai \
     -e POSTGRES_PASSWORD=ai \
     -e PGDATA=/var/lib/postgresql/data/pgdata \
     -v pgvolume:/var/lib/postgresql/data \
     -p 5532:5432 \
     --name pgvector \
     agnohq/pgvector:16
   

5. 运行代理

   Mac

   python cookbook/models/aws/claude/knowledge.py