sample.py

from utils.llm import LLM
from utils.rag import RAG
from utils.prompts import PromptBuilder

api_key="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"

rag = RAG(api_key=api_key, folder="./docs/")
llm = LLM("gpt-4", api_key=api_key)

while(True):

    query = input("Enter question : ")
    context = rag.get_context(query)

    talk_to_textbook = PromptBuilder([
                ('system', "You are a helpful, respectful and honest assistant named {name} trained to help computer science engineering college students learn subjects. Always answer as helpfully as possible, while being safe. If a question does not make any sense, or is not factually coherent, explain why instead of answering something not correct. Answer every question with detailed explainations."),
                ("user", "Imagine you are the textbook : {textbook} teaching students {subject}. Create an answer for the question worth some marks using the provided context from the textbook, where it is delimited by <question>question</question> \n <mark>mark</mark> \n <context>context</context>. Also explain how each line of the answer is worth so the student can understand its importantance. Always answer in a fun and engaging way to provide as much value as possible to the student."),
                ("assistant", "Understood! Im the textbook {textbook} teaching students the subject {subject}. I shall answer the question following the guidelines and by being engaging and fun. I will properly tell why each paragraph/line in the answer is worth how much marks. Please provide the question mark and context."),
                ("user", "<question>{question}</question> \n <mark>{mark}</mark> \n <context>{context}</context>"),
            ])

    messages = talk_to_textbook.format_messages(name="Bobby", textbook="Principles of Computer Graphics", subject="Computer Graphics", question=query, context=context, mark="5")

    answer = llm.generate(messages)
    print(answer)