qa_summary.py

import spaces
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

@spaces.GPU(duration=60)
def generate_answer(llm_name, texts, query, queries, response_lang, mode='validate'):
    
    if llm_name == 'solar':
        tokenizer = AutoTokenizer.from_pretrained("Upstage/SOLAR-10.7B-Instruct-v1.0", use_fast=True)
        llm_model = AutoModelForCausalLM.from_pretrained(
        "Upstage/SOLAR-10.7B-Instruct-v1.0",
        device_map="auto", #device_map="cuda"
        #torch_dtype=torch.float16,
        )

    elif llm_name == 'mistral':
        tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.2", use_fast=True)
        llm_model = AutoModelForCausalLM.from_pretrained(
        "mistralai/Mistral-7B-Instruct-v0.2",
        # device_map="auto", 
        device_map="cuda",
        torch_dtype=torch.float16,
        )

    elif llm_name == 'phi3mini':
        tokenizer = AutoTokenizer.from_pretrained("microsoft/Phi-3-mini-128k-instruct", use_fast=True)
        llm_model = AutoModelForCausalLM.from_pretrained(
        "microsoft/Phi-3-mini-128k-instruct", 
        device_map="auto", 
        torch_dtype="auto", 
        trust_remote_code=False,
        )  
    
    template_texts =""
    for i, text in enumerate(texts):
        template_texts += f'{i+1}. {text} \n'

    if mode == 'validate':
        conversation = [ {'role': 'user', 'content': f'Given the following query: "{query}"? \nIs the following document relevant to answer this query?\n{template_texts} \nResponse: Yes / No'} ]
    elif mode == 'summarize':
        conversation = [ {'role': 'user', 'content': f'For the following query and documents, try to answer the given query based on the documents.\nQuery: {query} \nDocuments: {template_texts}.'} ]
    elif mode == 'h_summarize':
        conversation = [ {'role': 'user', 'content': f'The documents below describe a developing disaster event. Based on these documents, write a brief summary in the form of a paragraph, highlighting the most crucial information. \nDocuments: {template_texts}'} ]
    elif mode == "multi_summarize":
        conversation = [ {'role': 'user', 'content': f"""For the following queries and documents, in a brief paragraph try to answer the given queries based on the documents. 
        Then, return the top 5 documents as provided that answer the queries.\nQueries: {queries} \nDocuments: {template_texts}. Give your response in {response_lang} language"""} ]
        

    prompt = tokenizer.apply_chat_template(conversation, tokenize=False, add_generation_prompt=True)
    inputs = tokenizer(prompt, return_tensors="pt").to(llm_model.device) 
    outputs = llm_model.generate(**inputs, use_cache=True, max_length=4096,do_sample=True,temperature=0.7,top_p=0.95,top_k=10,repetition_penalty=1.1)
    output_text = tokenizer.decode(outputs[0]) 
    if llm_name == "solar":
        assistant_respond = output_text.split("Assistant:")[1]
    elif llm_name == "phi3mini":
        assistant_respond = output_text.split("<|assistant|>")[1]
        assistant_respond = assistant_respond[:-7]
    else:
        assistant_respond = output_text.split("[/INST]")[1]
    if mode == 'validate':
        if 'Yes' in assistant_respond:
            return True
        else:
            return False
    elif mode == 'summarize':
        return assistant_respond
    elif mode == 'h_summarize':
        return assistant_respond
    elif mode == 'multi_summarize':
        return assistant_respond