Improve model card: Add pipeline tag, description, links, teaser image, and usage example

This PR significantly enhances the model card by:
- Adding the `pipeline_tag: unconditional-image-generation` for better discoverability on the Hugging Face Hub.
- Including a direct link to the research paper: [GAS: Improving Discretization of Diffusion ODEs via Generalized Adversarial Solver](https://huggingface.co/papers/2510.17699).
- Adding the official GitHub repository link: https://github.com/3145tttt/GAS.
- Incorporating the full abstract of the paper to provide a comprehensive description of the model.
- Adding the teaser image from the GitHub repository for visual context.
- Providing a practical sample usage code snippet from the GitHub README, demonstrating how to perform inference with a trained model.

Please review these additions and merge if everything looks good.

README.md

@@ -1,11 +1,41 @@
 ---
-license: mit
 datasets:
 - bayes-group-diffusion/GAS-teachers
+license: mit
 tags:
 - arxiv:2510.17699
+pipeline_tag: unconditional-image-generation
 ---
 
+# GAS: Improving Discretization of Diffusion ODEs via Generalized Adversarial Solver
+
+This repository contains the official implementation of the paper [GAS: Improving Discretization of Diffusion ODEs via Generalized Adversarial Solver](https://huggingface.co/papers/2510.17699).
+
+![Teaser image](https://github.com/3145tttt/GAS/raw/main/docs/teaser_1920.jpg)
+
+## Code
+The official code is available on GitHub: [https://github.com/3145tttt/GAS](https://github.com/3145tttt/GAS)
+
+## About
+While diffusion models achieve state-of-the-art generation quality, they still suffer from computationally expensive sampling. Recent works address this issue with gradient-based optimization methods that distill a few-step ODE diffusion solver from the full sampling process, reducing the number of function evaluations from dozens to just a few. However, these approaches often rely on intricate training techniques and do not explicitly focus on preserving fine-grained details. In this paper, we introduce the **Generalized Solver (GS)**: a simple parameterization of the ODE sampler that does not require additional training tricks and improves quality over existing approaches. We further combine the original distillation loss with adversarial training, which mitigates artifacts and enhances detail fidelity. We call the resulting method the **Generalized Adversarial Solver (GAS)** and demonstrate its superior performance compared to existing solver training methods under similar resource constraints.
+
+## Sample Usage
+To generate images from a trained **GS** or **GAS** checkpoint, you can use the `generate.py` script from the GitHub repository. First, ensure your environment is set up according to the instructions in the [GitHub README](https://github.com/3145tttt/GAS#setup-environment).
+
+Below is an example of generating 50,000 images on **CIFAR-10** using 2 GPUs and a checkpoint:
+
+```bash
+# Generate 50000 images using 2 GPUs and a checkpoint from checkpoint_path
+torchrun --standalone --nproc_per_node=2 generate.py \
+	--config=configs/edm/cifar10.yaml \
+	--outdir=data/teachers/cifar10 \
+	--seeds=50000-99999 \
+	--batch=1024 \
+	--steps=4 \
+	--checkpoint_path=checkpoint_path
+```
+For more detailed information on training, inference, and other configurations, please refer to the [official GitHub repository](https://github.com/3145tttt/GAS).
+
 ## Citation
 
 ```bibtex