Update app/core/model_loader.py

app/core/model_loader.py

@@ -9,14 +9,13 @@ from dotenv import load_dotenv
 from groq import Groq
 
 load_dotenv()
-
 logger = logging.getLogger(__name__)
-STATS_FILE = "usage_stats.json"
 
-# Supported active Groq free-tier models (Fallback list)
-DEFAULT_MODEL = "llama-3.1-8b-instant"
-FALLBACK_MODEL = "llama-3.3-70b-versatile"
+STATS_FILE = "usage_stats.json"
 
+# Updated active Groq production models matching current dashboard IDs
+DEFAULT_MODEL = "llama-3.1-8b"
+FALLBACK_MODEL = "llama-3.3-70b"
 
 class LLMSingleton:
     _instance = None
@@ -50,7 +49,6 @@ class LLMSingleton:
         raw_model = os.getenv("GROQ_MODEL", DEFAULT_MODEL)
         clean_model = raw_model.strip().strip('"').strip("'") if raw_model else ""
         self.model_name = clean_model if clean_model else DEFAULT_MODEL
-
         logger.info(f"🔑 Groq Client initialized with model target: {self.model_name}")
 
         # 3. Threading locks and usage stats setup
@@ -58,7 +56,6 @@ class LLMSingleton:
         self._rpm_lock = threading.Lock()
         self.stats = self._load_stats()
         self._check_daily_reset()
-
         self.rpm_limit = 30
         self.minute_window_start = time.time()
         self.requests_this_minute = 0
@@ -270,6 +267,5 @@ class LLMSingleton:
             logger.error(f"Groq text generation failed: {e}")
             return f"Error generating content: {str(e)}"
 
-
 # Export global instance
 llm_engine = LLMSingleton.get_instance()