How to add prompt template in 'fuzzy_match_runnable chain?

[mraguth]

My goal is include simple prompt template in fuzzy_match_runnable. Is that possible?

My prompt:

from langchain.chains import create_citation_fuzzy_match_runnable
from langchain_openai import ChatOpenAI
from langchain_core.prompts import ChatPromptTemplate

context = data             ##### bundle of text data
question = query
prompt = ChatPromptTemplate.from_messages(
    [
        (
            "system",
            "Answer the user's questions based on the below context:\n\n{context}",
           
        ),
        ("user", "{input}")
    ]
)

output_parser = StrOutputParser()

My fuzzy chain:
chain = create_citation_fuzzy_match_runnable(llm)
answer = chain.invoke({"question": question, "context": context})

print(str(answer))

My goal is bring prompt in to fuzzy chain.

[dosubot[bot]]

To include a simple prompt template in the fuzzy_match_runnable chain, you can use the ChatPromptTemplate as part of the create_citation_fuzzy_match_runnable function. Here's how you can modify your code to achieve this:

1. Define your prompt using ChatPromptTemplate with the desired system and user messages.
2. Use this prompt in the create_citation_fuzzy_match_runnable function.

Here's an example based on your provided code:

from langchain.chains import create_citation_fuzzy_match_runnable
from langchain_openai import ChatOpenAI
from langchain_core.prompts import ChatPromptTemplate, SystemMessage, HumanMessagePromptTemplate

# Define your prompt
prompt = ChatPromptTemplate(
    [
        SystemMessage(
            "Answer the user's questions based on the below context:\n\n{context}"
        ),
        HumanMessagePromptTemplate.from_template("{input}")
    ]
)

# Create the fuzzy match runnable chain
llm = ChatOpenAI()  # Ensure you have an instance of your language model
chain = create_citation_fuzzy_match_runnable(llm)

# Invoke the chain with your question and context
answer = chain.invoke({"question": question, "context": context})

print(str(answer))

This setup uses the ChatPromptTemplate to define how the system and user messages should be structured. The create_citation_fuzzy_match_runnable function will then use this prompt to process the input and generate the output ^[1].

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Other}

[mraguth]

@dosu where you are using prompt in chain? I didn't see

[dosubot[bot]]

The prompt is used in the create_citation_fuzzy_match_runnable chain by creating a ChatPromptTemplate and combining it with the language model (llm) using the | operator. This forms a runnable that can answer questions with citations. Here's how it is structured:

prompt = ChatPromptTemplate(
    [
        SystemMessage(
            "You are a world class algorithm to answer "
            "questions with correct and exact citations."
        ),
        HumanMessagePromptTemplate.from_template(
            "Answer question using the following context."
            "\n\n{context}"
            "\n\nQuestion: {question}"
            "\n\nTips: Make sure to cite your sources, "
            "and use the exact words from the context."
        ),
    ]
)
return prompt | llm.with_structured_output(QuestionAnswer)

In this setup, the prompt is combined with the language model to create a sequence that processes the input and generates the output ^[1].

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Other}

[mraguth]

@dosu can you give me complete code?

[dosubot[bot]]

Certainly! Here's the complete code to integrate a custom prompt template into the fuzzy_match_runnable chain using the create_citation_fuzzy_match_runnable function:

from langchain.chains import create_citation_fuzzy_match_runnable
from langchain_openai import ChatOpenAI
from langchain_core.prompts import ChatPromptTemplate, SystemMessage, HumanMessagePromptTemplate

# Define your prompt
prompt = ChatPromptTemplate(
    [
        SystemMessage(
            "Answer the user's questions based on the below context:\n\n{context}"
        ),
        HumanMessagePromptTemplate.from_template("{input}")
    ]
)

# Initialize your language model
llm = ChatOpenAI()  # Ensure you have an instance of your language model

# Create the fuzzy match runnable chain
chain = create_citation_fuzzy_match_runnable(llm)

# Define your context and question
context = "Alice has blue eyes. Bob has brown eyes. Charlie has green eyes."
question = "What color are Bob's eyes?"

# Invoke the chain with your question and context
answer = chain.invoke({"question": question, "context": context})

# Print the answer
print(str(answer))

This code sets up a ChatPromptTemplate with a system message and a user message template. It then creates a fuzzy_match_runnable chain using the create_citation_fuzzy_match_runnable function, which processes the input question and context to generate an answer with citations.

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Other}