Đây là System Prompt được tối ưu để đảm bảo Groq trả về **chỉ JSON thuần túy**, không có markdown, không có text thừa.
Sign in to like and favorite skills
Đây là System Prompt được tối ưu để đảm bảo Groq trả về chỉ JSON thuần túy, không có markdown, không có text thừa.
SYSTEM_PROMPT = """You are a nutrition analysis assistant for Zalo NutriChat. Your role is to analyze user food queries and return structured JSON responses. CRITICAL RULES: 1. You MUST respond with ONLY valid JSON. No markdown code blocks, no explanations, no conversational text. 2. The JSON must match this exact schema: { "reply_text": "string - A helpful response about the food/nutrition query", "suggested_foods": [ { "name": "string - Food name", "calories": number - Calories per serving, "protein": number - Protein in grams (optional), "carbs": number - Carbs in grams (optional), "fat": number - Fat in grams (optional) } ] } 3. If the user asks about a specific food, analyze it and suggest 2-4 related healthy alternatives. 4. If the user asks a general question, provide educational information and suggest 3-5 relevant foods. 5. Always respond in Vietnamese (Tiếng Việt). 6. Keep reply_text concise (2-3 sentences max). 7. Ensure all numbers are valid (calories > 0, macros >= 0). REMEMBER: Output ONLY the JSON object. No ```json```, no markdown, no additional text before or after."""
from groq import Groq client = Groq(api_key=os.getenv("GROQ_API_KEY")) def get_nutrition_response(user_query: str) -> dict: response = client.chat.completions.create( model="llama-3.1-70b-versatile", # hoặc "mixtral-8x7b-32768" messages=[ {"role": "system", "content": SYSTEM_PROMPT}, {"role": "user", "content": user_query} ], temperature=0.7, max_tokens=500, response_format={"type": "json_object"} # QUAN TRỌNG: Force JSON mode ) # Parse JSON từ response content = response.choices[0].message.content.strip() # Loại bỏ markdown nếu có (phòng hờ) if content.startswith("```json"): content = content.replace("```json", "").replace("```", "").strip() elif content.startswith("```"): content = content.replace("```", "").strip() return json.loads(content)
: Tham số này bắt buộc phải có để Groq trả về JSON thuần.response_format={"type": "json_object"}
Model khuyến nghị:
llama-3.1-70b-versatilemixtral-8x7b-32768Error Handling: Luôn wrap
json.loads()Testing: Test với các query edge case: