test: latency test with lstm model (transformers latency is 450-950ms excluding overhead)

inference_api.py

@@ -35,7 +35,7 @@ logger = logging.getLogger("DKTInferenceEngine")
 # Global Configuration
 # ──────────────────────────────────────────────────────────────
 # Easy Switch Model Choice: 'LSTM' or 'Transformer'
-MODEL_TYPE = "Transformer"
+MODEL_TYPE = "LSTM"
 
 # Thresholds
 STRUGGLE_THRESHOLD = 0.50     # Skills below this are flagged as struggling