Update README.md

README.md

@@ -15,6 +15,9 @@ This is a int4_awq quantized checkpoint of bigcode/starcoder2-15b. It takes abou
 ## Running this Model
 vLLM does not natively support autoawq currently (or any a4w8 as of writing this), so one can just serve directly from the autoawq backend.
 
+Note, if you want to start this in a container, then:
+`docker run --gpus all -it --name=starcoder2-15b-int4-awq -p 8000:8000 -v ~/.cache:/root/.cache nvcr.io/nvidia/pytorch:24.12-py3 bash`
+
 `pip install fastapi[all] torch transformers autoawq`
 
 Then in python3: