#!pip install langchain openai pymilvus ragas

OPENAI_API_KEY="<YOUR_OPENAI_API_KEY>"

import requestsfrom langchain.document_loaders import TextLoaderfrom langchain.text_splitter import CharacterTextSplitter
url = "https://publicdataset.zillizcloud.com/milvus_doc.md"res = requests.get(url)with open("milvus_doc.md", "w") as f:    f.write(res.text)
# Load the dataloader = TextLoader('./milvus_doc.md')documents = loader.load()
# Chunk the datatext_splitter = CharacterTextSplitter(chunk_size=60, chunk_overlap=10)chunks = text_splitter.split_documents(documents)

from langchain.embeddings import OpenAIEmbeddingsfrom langchain.vectorstores.milvus import Milvusvectorstore = Milvus.from_documents(    chunks,    embedding=OpenAIEmbeddings(),    connection_args={"host": "127.0.0.1", "port": "19530"},)retriever = vectorstore.as_retriever(search_kwargs={"k": 3})

from langchain.chat_models import ChatOpenAIfrom langchain.prompts import ChatPromptTemplatefrom langchain.schema.runnable import RunnablePassthroughfrom langchain.schema.output_parser import StrOutputParser
# Define LLMllm = ChatOpenAI(model_name="gpt-3.5-turbo", temperature=0)
# Define prompt templatetemplate = """You are an assistant for question-answering tasks. Use the following pieces of retrieved context to answer the question. If you don't know the answer, just say that you don't know. Use two sentences maximum and keep the answer concise.Question: {question} Context: {context} Answer:"""
prompt = ChatPromptTemplate.from_template(template)
# Setup RAG pipelinerag_chain = (    {"context": retriever,  "question": RunnablePassthrough()}     | prompt     | llm    | StrOutputParser() )

from datasets import Dataset
questions = ["Can user delete milvus entities through non-primary key filtering?"            ]ground_truths = [["Milvus supports deleting entities by primary key or complex boolean expressions"]]answers = []contexts = []
# Inferencefor query in questions:    answers.append(rag_chain.invoke(query))    contexts.append([docs.page_content for docs in retriever.get_relevant_documents(query)])
data = {    "question": questions,    "answer": answers,    "contexts": contexts,    "ground_truths": ground_truths}
# Convert dict to datasetdataset = Dataset.from_dict(data)

from ragas import evaluatefrom ragas.metrics import (    faithfulness,    answer_relevancy,    context_recall,    context_precision,)result = evaluate(    dataset = dataset,    metrics=[        context_precision,        context_recall,        faithfulness,        answer_relevancy,    ],)df = result.to_pandas()