force kill vllm instance to free gpu resources

Author: aarora79

src/fmbench/scripts/inference_containers/vllm_gpu.py

@@ -15,8 +15,16 @@ def create_script(region, image_uri, model_id, model_name, env_str, privileged_s
     script = f"""#!/bin/sh
 
         {STOP_AND_RM_CONTAINER}
-
-        
+        # kill any existing vllm instance to free up gpu resources
+        process_name="vllm"
+        pid=$(pgrep -f "$process_name")
+        if [ -n "$pid" ]; then
+          echo "Killing process $process_name with PID $pid"
+          kill -9 "$pid"
+        else
+          echo "No process named $process_name is running"
+        fi
+        sleep 10
         vllm serve {model_id}
 
         echo "started docker run in daemon mode"