使用 Qdrant、CAMEL-AI 和 OpenAI 构建基于 Agentic RAG 的 Discord 聊天机器人

!pip install camel-ai[all]==0.2.17

OPENAI_API_KEY=<your_openai_api_key>

import openai
import os
from dotenv import load_dotenv

load_dotenv()

openai_client = openai.Client(
    api_key=os.getenv("OPENAI_API_KEY")
)

from camel.embeddings import OpenAIEmbedding
from camel.types import EmbeddingModelType

embedding_instance = OpenAIEmbedding(model_type=EmbeddingModelType.TEXT_EMBEDDING_3_LARGE)

QDRANT_CLOUD_URL=<your-qdrant-cloud-url>
QDRANT_CLOUD_API_KEY=<your-api-key>

from camel.retrievers import VectorRetriever

# Define collection name
collection_name = "qdrant-agent"

storage_instance = QdrantStorage(
    vector_dim=embedding_instance.get_output_dim(),
    url_and_api_key=(
        qdrant_cloud_url,
        qdrant_api_key,
    ),
    collection_name=collection_name,
)

from camel.retrievers import VectorRetriever

vector_retriever = VectorRetriever(embedding_model=embedding_instance,
                                   storage=storage_instance)

qdrant_urls = [
    "https://qdrant.org.cn/documentation/overview",
    "https://qdrant.org.cn/documentation/guides/installation",
    "https://qdrant.org.cn/documentation/concepts/filtering",
    "https://qdrant.org.cn/documentation/concepts/indexing",
    "https://qdrant.org.cn/documentation/guides/distributed_deployment",
    "https://qdrant.org.cn/documentation/guides/quantization"
    # Add more URLs as needed
]

for qdrant_url in qdrant_urls:
  vector_retriever.process(
      content=qdrant_url,
  )

from camel.configs import ChatGPTConfig
from camel.models import ModelFactory
from camel.types import ModelPlatformType, ModelType
from camel.agents import ChatAgent

# Create a ChatGPT configuration
config = ChatGPTConfig(temperature=0.2).as_dict()

# Create an OpenAI model using the configuration
openai_model = ModelFactory.create(
    model_platform=ModelPlatformType.OPENAI,
    model_type=ModelType.GPT_4O_MINI,
    model_config_dict=config,
)

assistant_sys_msg = """You are a helpful assistant to answer question,
         I will give you the Original Query and Retrieved Context,
        answer the Original Query based on the Retrieved Context,
        if you can't answer the question just say I don't know."""

qdrant_agent = ChatAgent(system_message=assistant_sys_msg, model=openai_model)

DISCORD_BOT_TOKEN=<your-discord-bot-token>

from camel.bots import DiscordApp
import nest_asyncio
import discord

nest_asyncio.apply()
discord_q_bot = DiscordApp(token=os.getenv("DISCORD_BOT_TOKEN"))

@discord_q_bot.client.event # triggers when a message is sent in the channel
async def on_message(message: discord.Message):
    if message.author == discord_q_bot.client.user:
        return

    if message.type != discord.MessageType.default:
        return

    if message.author.bot:
        return
    user_input = message.content

    retrieved_info = vector_retriever.query(
        query=user_input, top_k=10, similarity_threshold=0.6
    )

    user_msg = str(retrieved_info)
    assistant_response = qdrant_agent.step(user_msg)
    response_content = assistant_response.msgs[0].content

    if len(response_content) > 2000: # discord message length limit
        for chunk in [response_content[i:i+2000] for i in range(0, len(response_content), 2000)]:
            await message.channel.send(chunk)
    else:
        await message.channel.send(response_content)

discord_q_bot.run()