app.py

import streamlit as st
import transformers
import torch
import os
from transformers import AutoTokenizer, AutoModelForCausalLM
from huggingface_hub import login

# Retrieve the Hugging Face token from the environment variables
HUGGINGFACE_TOKEN = os.getenv("HUGGINGFACE_TOKEN")
if HUGGINGFACE_TOKEN is None:
    st.error("Hugging Face token is not set. Please add it to the environment variables.")
    st.stop()

# Authenticate using the token
login(HUGGINGFACE_TOKEN)

model_id = "meta-llama/Meta-Llama-3-8B"

# Load model and tokenizer
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=torch.bfloat16)

generator = transformers.pipeline(
    "text-generation", model=model, tokenizer=tokenizer, device_map="auto"
)

text = st.text_area("Ask me Anything")

if text:
    result = generator(text)
    st.success(result)