app.py

import torch
from transformers import AutoTokenizer, AutoModelForCausalLM
import gradio as gr

model_id = "mistralai/Mistral-7B-Instruct-v0.1"

tokenizer = AutoTokenizer.from_pretrained(model_id, use_auth_token=True)
model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=torch.float16, device_map="auto")

chat_history = []

def chat(user_input):
    global chat_history

    messages = [{"role": "system", "content": "You are a helpful assistant."}]
    for message in chat_history:
        messages.append({"role": "user", "content": message[0]})
        messages.append({"role": "assistant", "content": message[1]})
    messages.append({"role": "user", "content": user_input})

    prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)

    output = model.generate(**inputs, max_new_tokens=200, do_sample=True, temperature=0.7)
    decoded_output = tokenizer.decode(output[0], skip_special_tokens=True)
    answer = decoded_output.split(messages[-1]["content"])[-1].strip()

    chat_history.append((user_input, answer))
    return answer

iface = gr.Interface(fn=chat, inputs="text", outputs="text", title="Mistral 7B Chat")
iface.launch()