Update README.md

README.md

@@ -42,6 +42,10 @@ with torch.no_grad():
     probability = torch.softmax(logits, dim=1).detach().numpy()[:,1].tolist() #  Probability of toxicity.
    
 ```
+
+## Cookbook on Model Usage as a Guardrail
+This recipe illustrates the use of the model either in a prompt, the output, or both. This is an example of a “guard rail” typically used in generative AI applications for safety.
+[Guardrail Cookbook](https://github.com/ibm-granite-community/granite-code-cookbook/blob/main/recipes/Guard-Rails/HAP.ipynb)
 ## Performance Comparison with Other Models
 The model outperforms most popular models with significantly lower inference latency. If a better F1 score is required, please refer to IBM's 12-layer model [here](https://huggingface.co/ibm-granite/granite-guardian-hap-125m).