app.py

import gradio as gr
from datasets import load_dataset
import os
from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer, BitsAndBytesConfig
import torch
from threading import Thread
from sentence_transformers import SentenceTransformer
from datasets import load_dataset
import time

token = os.environ["HF_TOKEN"]

ST = SentenceTransformer("mixedbread-ai/mxbai-embed-large-v1")
art_dataset= load_dataset("hichri-mo/arxiver-1000",revision="embedded")

data = art_dataset["train"]
data = data.add_faiss_index("embeddings")


model_id= "Qwen/Qwen2.5-3B-Instruct"

# use quantization to lower GPU usage
bnb_config = BitsAndBytesConfig(
    load_in_4bit=True, bnb_4bit_use_double_quant=True, bnb_4bit_quant_type="nf4", bnb_4bit_compute_dtype=torch.bfloat16
)

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    torch_dtype=torch.bfloat16,
    device_map="auto",
    quantization_config=bnb_config
)
terminators = [
    tokenizer.eos_token_id,
    tokenizer.convert_tokens_to_ids("<|eot_id|>")
]

SYS_PROMPT = """You are an assistant for answering questions.
You are given the extracted parts of a long document and a question. Provide a conversational answer.
If you don't know the answer, just say "I do not know." Don't make up an answer."""


def format_prompt(prompt,retrieved_documents,k):
  """using the retrieved documents we will prompt the model to generate our responses"""
  PROMPT = f"Question: {prompt}\nContext: \n"
  for idx in range(k) :
    PROMPT+= f"{retrieved_documents['markdown'][idx]}\n"
  return PROMPT

def generate(formatted_prompt):
  formatted_prompt = formatted_prompt[:2000] # to avoid GPU OOM
  messages = [{"role":"system","content":SYS_PROMPT},{"role":"user","content":formatted_prompt}]
  # tell the model to generate
  input_ids = tokenizer.apply_chat_template(
      messages,
      add_generation_prompt=True,
      return_tensors="pt"
  ).to(model.device)
  # Check if terminators contain None and replace with tokenizer.eos_token_id
  eos_token_id = terminators[0]  # Default to tokenizer.eos_token_id
  if terminators[1] is not None:
    eos_token_id = terminators[1]  # Use "<|eot_id|>" if it exists
  
  outputs = model.generate(
      input_ids,
      max_new_tokens=1024,
      eos_token_id=eos_token_id, # Pass a single integer value
      do_sample=True,
      temperature=0.6,
      top_p=0.9,
  )
  response = outputs[0][input_ids.shape[-1]:]
  return tokenizer.decode(response, skip_special_tokens=True)

def rag_chatbot(prompt:str,k:int=2):
  scores , retrieved_documents = search(prompt, k)
  formatted_prompt = format_prompt(prompt,retrieved_documents,k)
  return generate(formatted_prompt)




def rag_chatbot_interface(prompt, k):
  return rag_chatbot(prompt, k)

iface = gr.Interface(
    fn=rag_chatbot_interface,
    inputs=[
        gr.Textbox(label="Enter your question"),
        gr.Slider(minimum=1, maximum=10, step=1, value=2, label="Number of documents to retrieve")
    ],
    outputs=gr.Textbox(label="Response"),
    title="Chatbot with RAG",
    description="Ask questions and get answers based on retrieved documents."
)

iface.launch()