import requests # URL de ton endpoint url = "https://micksoftware-laria-startup.hf.space/predict" # Données pour la requête payload = { "message": "quelle longueur d'hypothenuse pour un triangle de cote 4 et 9", "history": [], "system_message": "You are a friendly assistant. Answer questions briefly and do not continue conversations unless explicitly asked.", "max_tokens": 512, "temperature": 0.4, "top_p": 0.95, } # Faire la requête avec streaming activé response = requests.post(url, json=payload, stream=True) # Lire la réponse en streaming if response.status_code == 200: print("Streaming response:") for chunk in response.iter_lines(decode_unicode=True): if chunk: print(chunk, end="") else: print(f"Erreur : {response.status_code} - {response.text}")