Update start_gradio.sh

start_gradio.sh

@@ -89,4 +89,17 @@ if [[ "$PRELOAD_MODEL" == "1" ]]; then
         echo "✗ model.safetensors.index.json 缺失"
     fi
     echo "==================="
-fi
+fi
+
+echo "模型路径: ${MODEL_PATH}"
+echo "启动 Gradio..."
+
+python app.py \
+    --host "$HOST" \
+    --port "$GRADIO_PORT" \
+    --model "$MODEL_PATH" \
+    --tensor-parallel-size "$TENSOR_PARALLEL_SIZE" \
+    --max-model-len "$MAX_MODEL_LEN" \
+    --gpu-memory-utilization "$GPU_MEMORY_UTILIZATION" \
+    --tokenizer-mode "$TOKENIZER_MODE" \
+    --served-model-name "$SERVED_MODEL_NAME"