src/llm/chain.py

from langchain.prompts import ChatPromptTemplate
from langchain.schema.runnable import RunnableLambda
from operator import itemgetter
from langchain.output_parsers import PydanticOutputParser
from .output_parser import SongDescriptions
from langchain.llms.base import LLM


class LLMChain:
    def __init__(self, llm_model: LLM) -> None:
        self.llm_model = llm_model
        self.parser = PydanticOutputParser(pydantic_object=SongDescriptions)
        self.full_chain = self._create_llm_chain()


    def _get_output_format(self, _):
        return self.parser.get_format_instructions()
        
    def _create_llm_chain(self):
        prompt_response = ChatPromptTemplate.from_messages([
            ("system", "You are an AI assistant, helping the user to turn a music playlist text description into four separate song descriptions that are probably contained in the playlist. Try to be specific with descriptions. Make sure all 4 song descriptions are similar.\n"),
            ("system", "{format_instructions}\n"),
            ("human", "Playlist description: {description}.\n"),
            # ("human", "Song descriptions:"),
        ])
        # prompt = PromptTemplate(
        #     template="You are an AI assistant, helping the user to turn a music playlist text description into three separate generic song descriptions that are probably contained in the playlist.\n{format_instructions}\n{description}\n",
        #     input_variables=["description"],
        #     partial_variables={"format_instructions": self.parser.get_format_instructions()},
        # )


        full_chain = (
            {
                "format_instructions": RunnableLambda(self._get_output_format),
                "description": itemgetter("description"),
            } 
            | prompt_response
            | self.llm_model
        )
        return full_chain
    
    def process_user_description(self, user_input):
        output = self.full_chain.invoke(
            {
                "description": user_input
            }
        ).replace("\\", '')
        return self.parser.parse(output)