Update start.sh

start.sh

@@ -1,6 +1,7 @@
 #!/bin/bash
 # Start vLLM server in the background with your custom flags
 vllm \
+  --runtime nvidia --gpus all --ipc=host \
   --model unsloth/llama-3-8b-Instruct-bnb-4bit \
   --enable-auto-tool-choice \
   --tool-call-parser llama3_json \