app.py

from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

def main():
    model_id = "meta-llama/Llama-3.1-8B"

    print("Chargement du tokenizer...")
    tokenizer = AutoTokenizer.from_pretrained(model_id)

    print("Chargement du modèle...")
    model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=torch.float16, device_map="auto")

    prompt = "Bonjour, je suis une IA super intelligente appelée io,"

    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)

    print("Génération du texte...")
    outputs = model.generate(
        **inputs,
        max_length=100,
        do_sample=True,
        temperature=0.7,
        top_p=0.9,
        num_return_sequences=1
    )

    text = tokenizer.decode(outputs[0], skip_special_tokens=True)
    print("\nTexte généré :\n", text)

if __name__ == "__main__":
    main()