Update README.md file

README.md

@@ -1,13 +1,106 @@
----
-license: mit
-task_categories:
-- reinforcement-learning
-language:
-- en
-tags:
-- multi-agent
-- vehicle-routing
----
-Multi Agent Environments for Vehicle Routing Problems
-
-MAEnvs4VRP is a library made up of multi-agent environments for simulating classic vehicle routing problems.
+---
+language:
+- en
+license: mit
+pretty_name: MAEnvs4VRP Datasets
+tags:
+- reinforcement-learning
+- combinatorial-optimization
+- vehicle-routing-problem
+- operations-research
+- multi-agent
+- dataset
+task_categories:
+- reinforcement-learning
+size_categories:
+- unknown
+---
+
+<p align="center">
+<img src="https://raw.githubusercontent.com/MAEnvs4VRP/maenvs4vrp/master/docs/source/img/banner.png" width="600" />
+</p>
+
+Main code on Github: https://github.com/MAEnvs4VRP/maenvs4vrp
+
+
+## MAEnvs4VRP Datasets
+
+The **MAEnvs4VRP Datasets** repository provides benchmark and synthetic datasets for the vehicle routing problems (VRPs) supported by the **MAEnvs4VRP** framework.
+
+The repository is intended to serve as a centralized collection of routing instances for training, validating, and benchmarking reinforcement learning algorithms and other optimization methods on classical and advanced Vehicle Routing Problems.
+
+The datasets are fully compatible with the MAEnvs4VRP library and can be used to reproduce experiments, evaluate new learning algorithms, and compare optimization approaches across multiple VRP variants.
+
+---
+
+# Supported Vehicle Routing Problems
+
+The repository contains datasets for several routing problems implemented in MAEnvs4VRP, including:
+
+| Problem | Description |
+|----------|-------------|
+| CVRP | Capacitated Vehicle Routing Problem |
+| CVRPSTW | Capacitated Vehicle Routing Problem with Time Windows |
+| CVRPTW | Capacitated Vehicle Routing Problem with Soft Time Windows |
+| PDPTW | Pickup and Delivery Problem with Time Windows |
+| SDVRPTW | Split Delivery Vehicle Routing Problem with Time Windows |
+| TOP | Team Orienteering Problem |
+| TOPTW | Team Orienteering Problem with Time Windows |
+
+---
+
+# Using the Datasets
+
+The datasets provided in this repository are designed to be used with the **MAEnvs4VRP** framework.
+
+Benchmark instances are automatically downloaded from the Hugging Face Hub when required by the corresponding environment. After the initial download, the datasets are cached locally and reused in future executions.
+
+Each environment in MAEnvs4VRP provides the appropriate instance generator for loading and sampling benchmark instances. Refer to the documentation of the corresponding environment for usage examples and configuration options.
+
+---
+
+# Relationship with MAEnvs4VRP
+
+This repository complements the **MAEnvs4VRP** library by providing the datasets used by its environments.
+
+The library implements simulation environments for multiple Vehicle Routing Problems, while this repository stores the corresponding benchmark and synthetic routing instances.
+
+---
+
+# Limitations
+
+Users should be aware that:
+
+- different routing problems contain different data schemas;
+- not every environment currently includes the same number of benchmark instances;
+- additional datasets may be incorporated in future releases;
+- synthetic datasets are generated according to the assumptions of their corresponding environment.
+
+---
+
+# Citation
+
+If you use these datasets in your research, please cite the MAEnvs4VRP paper.
+
+```bibtex
+@article{maenvs4vrp,
+  title={MAEnvs4VRP: Multi-Agent Environments for Vehicle Routing Problems},
+  author={}
+  year={2026},
+  journal={INFORMS Journal on Computing},
+}
+```
+
+---
+
+# License
+
+This dataset is distributed under the MIT License unless otherwise specified for individual benchmark collections.
+
+Users should also acknowledge the original authors of benchmark datasets whenever applicable.
+
+---
+
+# Acknowledgements
+
+This repository is part of the **MAEnvs4VRP** project, which provides a unified framework for developing and evaluating reinforcement learning algorithms for Vehicle Routing Problems.