app/chains.py

import os
from langchain_huggingface import HuggingFaceEndpoint
from langchain_core.runnables import RunnablePassthrough
from transformers import AutoTokenizer
model_id = "meta-llama/Llama-3.2-3B-Instruct"
tokenizer = AutoTokenizer.from_pretrained(model_id)

import schemas
from prompts import (
    raw_prompt,
    raw_prompt_formatted, 
    history_prompt_formatted, 
    standalone_prompt_formatted,
    rag_prompt_formatted,
    format_context,
    tokenizer
)
from data_indexing import DataIndexer
data_indexer = DataIndexer()

llm = HuggingFaceEndpoint(
    repo_id=model_id
    huggingfacehub_api_token=os.environ['HF_TOKEN'],
    max_new_tokens=512,
    stop_sequences=[tokenizer.eos_token],
    streaming=True,
)

simple_chain = (raw_prompt | llm).with_types(input_type=schemas.UserQuestion)

formatted_chain = (
    raw_prompt_formatted
    | llm
).with_types(input_type=schemas.UserQuestion)

history_chain = (
    history_prompt_formatted
    | llm
).with_types(input_type=schemas.HistoryInput)

standalone_prompt_formatted =
format_prompt(standalone_prompt)
standalone_chain = standalone_prompt_formatted | llm

generation_chain = rag_prompt_formatted | llm

rag_chain = (
    RunnablePassthrough.assign(new_question=standalone_chain)
    | {
        'context': lambda x:
format_context(search(x['new_question'])),
        'standalone_question': lambda x: x['new_question']
    }
    | generation_chain
).with_types(input_type=schemas.RagInput)