Update start.sh

start.sh

@@ -36,22 +36,20 @@ except Exception as e:
 # 1. TEXT: Qwen 2.5 Coder
 download_model "Qwen/Qwen2.5-Coder-14B-Instruct-GGUF" "qwen2.5-coder-14b-instruct-q4_k_m.gguf"
 
-# 2. IMAGE: Qwen 2.5 VL (FIXED: correct case-sensitive filename)
+# 2. IMAGE: Qwen 2.5 VL (correct case-sensitive filename)
 download_model "ggml-org/Qwen2.5-VL-7B-Instruct-GGUF" "Qwen2.5-VL-7B-Instruct-Q4_K_M.gguf" "qwen2.5-vl-7b-instruct-q4_k_m.gguf"
 
 # 3. AUDIO: Qwen 2 Audio
 download_model "gaianet/Qwen2-Audio-7B-Instruct-GGUF" "Qwen2-Audio-7B-Instruct-Q4_K_M.gguf" "qwen2-audio-7b-instruct-q4_k_m.gguf"
 
 echo "------------------------------------------------"
-echo "🚀 Starting Server"
+echo "🚀 Starting Server in Router Mode"
 echo "------------------------------------------------"
 
-MODEL_FILE="/app/models/qwen2.5-coder-14b-instruct-q4_k_m.gguf"
-
-echo "Selected Model: $MODEL_FILE"
-
+# Start in ROUTER MODE - serve all models from /app/models
 exec /usr/local/bin/llama-server \
-    -m "$MODEL_FILE" \
+    --models-dir /app/models \
+    --models-max 3 \
     --host 0.0.0.0 \
     --port 7860 \
     -c 8192 \