LLM驱动节点生成聊天内容

fastapi作为后端框架。MySQL数据持久化。 ollama本地LLM，llama3.1-8b

创建config.py文件，输入

MYSQLPASSWORD = 'YOUR_PASSWORD'
SQLALCHEMY_DATABASE_URL = "mysql+pymysql://root:" + MYSQLPASSWORD + "@localhost/social_network_simulation"

curl -X POST "http://localhost:8000/nodes/" -H "Content-Type: application/json" -d '{"name": "Amy"}'

curl -X POST "http://localhost:8000/nodes/1/connect" \ -H "Content-Type: application/json" \ -d '{"target_node_id": 2}'

curl -X POST "http://localhost:8000/start-dialogue/?node1_id=1&node2_id=2&num_turns=5"

curl -X POST "http://localhost:8000/multi-dialogue/?num_turns=5"

可视化 从真实数据中提取节点profile

• https://dl.acm.org/doi/abs/10.1145/3586183.3606763
• https://arxiv.org/abs/2307.14984
• https://arxiv.org/abs/2311.09618