Update model card with paper link, graph-ml tag, and usage instructions

README.md

@@ -1,16 +1,48 @@
 ---
-license: mit
 datasets:
 - ai4co/rrnco
-pipeline_tag: reinforcement-learning
+license: mit
+pipeline_tag: graph-ml
 ---
 
-# RRNCO models
+# RRNCO: Towards Real-World Routing with Neural Combinatorial Optimization
+
+This repository contains the pre-trained models for **RRNCO**, as presented in the paper [RRNCO: Towards Real-World Routing with Neural Combinatorial Optimization](https://huggingface.co/papers/2503.16159).
+
+RRNCO is a novel architecture designed to bridge the sim-to-real gap in Neural Combinatorial Optimization (NCO) for Vehicle Routing Problems (VRPs). It introduces two key innovations: **Adaptive Node Embedding (ANE)** and **Neural Adaptive Bias (NAB)** to handle complex real-world features like asymmetric distance and duration matrices.
+
+- **Main code on Github:** [ai4co/real-routing-nco](https://github.com/ai4co/real-routing-nco)
+- **Dataset:** [ai4co/rrnco](https://huggingface.co/datasets/ai4co/rrnco)
+
+## Usage
+
+You can evaluate the pre-trained checkpoints (available under the [`checkpoints/`](https://huggingface.co/ai4co/rrnco/tree/main/checkpoints) folder) using the `test.py` script from the official repository.
+
+### ATSP (Asymmetric Traveling Salesman Problem)
+```bash
+python test.py --problem atsp --datasets data/atsp/atsp_n100_seed3333_in_distribution.npz --batch_size 32 --checkpoint checkpoints/atsp/epoch_199.ckpt
+```
 
+### RCVRP (Real-world Capacitated VRP)
+```bash
+python test.py --problem rcvrp --datasets data/rcvrp/rcvrp_n100_seed3333_in_distribution.npz --batch_size 32 --checkpoint checkpoints/rcvrp/epoch_199.ckpt
+```
 
-These are the models accompanying our work ["Neural Combinatorial Optimization for Real-World Routing"](https://arxiv.org/abs/2503.16159)
+### RCVRPTW (Real-world Capacitated VRP with Time Windows)
+```bash
+python test.py --problem rcvrptw --datasets data/rcvrptw/rcvrptw_n100_seed3333_in_distribution.npz --batch_size 32 --checkpoint checkpoints/rcvrptw/epoch_199.ckpt
+```
 
+## Citation
 
-Main code on Github: https://github.com/ai4co/real-routing-nco
+If you find RRNCO valuable for your research, please cite the work:
 
-Models are available under the [`checkpoints/`](https://huggingface.co/ai4co/rrnco/tree/main/checkpoints) folder
+```bibtex
+@inproceedings{son2026rrnconeuralcombinatorialoptimizationrealworldrouting,
+  title     = {Towards Real-World Routing with Neural Combinatorial Optimization},
+  author    = {Son, Jiwoo and Zhao, Zhikai and Berto, Federico and Hua, Chuanbo and Cao, Zhiguang and Kwon, Changhyun and Park, Jinkyoo},
+  booktitle = {Proceedings of the International Conference on Learning Representations (ICLR)},
+  year      = {2026},
+  url       = {https://openreview.net/forum?id=sKvo9ZZfpe}
+}
+```