Add model card for DiffGap

This PR improves the model card for DiffGap by:
- Adding the `pipeline_tag: other` metadata.
- Linking the model to the associated research paper: [Bridging the Gap between Learning and Inference for Diffusion-Based Molecule Generation](https://huggingface.co/papers/2411.05472).
- Adding a link to the official GitHub repository.
- Providing sample usage instructions for training, sampling, and evaluation based on the official implementation.
- Including the BibTeX citation for the paper.

README.md

@@ -1,9 +1,55 @@
+---
+pipeline_tag: other
+---
+
 # DiffGap
 
-The checkpoints and metrics files for [DiffGap](https://github.com/HUGHNew/DiffGap).
+This repository contains the official checkpoints and metrics for **DiffGap**, as introduced in the paper [Bridging the Gap between Learning and Inference for Diffusion-Based Molecule Generation](https://huggingface.co/papers/2411.05472).
+
+DiffGap is a diffusion-based framework that integrates adaptive sampling and pseudo-molecule estimation to bridge the gap between training objectives and inference dynamics in 3D molecule generation. By dynamically aligning intermediate denoising steps with realistic generation trajectories, DiffGap enables the diffusion model to adapt to input biases in advance during the training phase.
+
+- **Repository:** [https://github.com/neusymlab/DiffGap](https://github.com/neusymlab/DiffGap)
+- **Paper:** [https://huggingface.co/papers/2411.05472](https://huggingface.co/papers/2411.05472)
+
+## Repository Structure
+
+The repository includes:
+- **ckpts**: Baseline checkpoints and DiffGap model weights.
+- **metrics**: Sampling and evaluation results.
+
+## Usage
+
+The official implementation uses a `pipeline.py` script to wrap the processes of training, sampling, and evaluation.
+
+```bash
+# Basic command structure
+python -m pipeline <configs> <sampling_results> [train|sample|eval] [-c resume_from_checkpoint_for_training]
+
+# Example: Run the whole pipeline
+# python -m pipeline configs/training.yml sampling_results/reproduce 
+
+# Example: Run sampling from a specific config
+# python -m pipeline configs/sampling.yml sampling_results/reproduce sample 
+
+# Example: Run evaluation
+# python -m pipeline "no matter" sampling_results/reproduce eval 
+```
+
+For sampling on the PDBbind dataset with the DiffGap configuration:
+```bash
+python pipeline.py configs/gbd_pdbbind.yaml sampling_results/binddm_pdbbind sample -p PDBbind_refined_2020_test
+```
 
-Including files:
-- ckpts: baseline checkpoints and ours
-- metrics: the sampling and evaluation results
+## Citation
 
-We list all baselines for convenience.
+```bibtex
+@misc{liu2024gapdiff,
+      title={Bridging the Gap between Learning and Inference for Diffusion-Based Molecule Generation}, 
+      author={Peidong Liu and Wenbo Zhang and Xue Zhe and Jiancheng Lv and Xianggen Liu},
+      year={2024},
+      eprint={2411.05472},
+      archivePrefix={arXiv},
+      primaryClass={cs.LG},
+      url={https://arxiv.org/abs/2411.05472}, 
+}
+```