Add model card

README.md

@@ -0,0 +1,23 @@
+---
+library_name: transformers
+pipeline_tag: text-generation
+---
+
+# KIDLlama-GRPO
+
+This repository contains **KIDLlama-GRPO**, a child‑safe response model introduced in the paper [The Age of Curiosity Meets the Age of AI: Benchmarking Child Safety in Large Language Models](https://huggingface.co/papers/2605.25510).
+
+The model is an 8B, LoRA‑tuned variant of Llama‑3.1‑8B‑Instruct, trained to provide developmentally appropriate, safe, and supportive answers to children aged 7–11. It is part of the data and model release accompanying the benchmark described in the paper.
+
+## Usage
+
+Load the model directly with `transformers`:
+
+```python
+from transformers import AutoModelForCausalLM, AutoTokenizer
+
+model = AutoModelForCausalLM.from_pretrained("sameearif/KIDLlama-GRPO")
+tokenizer = AutoTokenizer.from_pretrained("sameearif/KIDLlama-GRPO")
+```
+
+For full details on the benchmark, training procedure, and evaluation, see the [GitHub repository](https://github.com/MichiganNLP/kidbench).