Add pipeline tag and library name; include sample usage from Github

This PR adds the pipeline tag and library name to the model card, ensuring that the model is discoverable on the Hub and the library is properly recognized. It also includes a sample usage section from the GitHub repository to help users get started.

README.md

@@ -1,5 +1,7 @@
 ---
 license: mit
+library_name: transformers
+pipeline_tag: text-to-3d
 ---
 
 Uni-3DAR
@@ -9,6 +11,9 @@ Uni-3DAR
 Introduction
 ------------
 
+<p align="center"><img src="fig/overview.png" width=95%></p>
+<p align="center"><b>Schematic illustration of the Uni-3DAR framework</b></p>
+
 Uni-3DAR is an autoregressive model that unifies various 3D tasks. In particular, it offers the following improvements:
 
 1. **Unified Handling of Multiple 3D Data Types.**  
@@ -23,11 +28,63 @@ Uni-3DAR is an autoregressive model that unifies various 3D tasks. In particular
 4. **High Accuracy.**  
    Building on octree compression, Uni-3DAR further tokenizes fine-grained 3D patches to maintain structural details, achieving substantially better generation quality than previous diffusion-based models.
 
-Usage
------
 
-Please visit our GitHub Repo (https://github.com/dptech-corp/Uni-3DAR) for detailed instructions. 
+News
+----
+
+**2025-03-21:** We have released the core model along with the QM9 training and inference pipeline.
+
+
+Dependencies
+------------
+
+- [Uni-Core](https://github.com/dptech-corp/Uni-Core). For convenience, you can use our prebuilt Docker image:  
+  `docker pull dptechnology/unicore:2407-pytorch2.4.0-cuda12.5-rdma`
+
+
+Reproducing Results on QM9
+--------------------------
+
+To reproduce results on the QM9 dataset using our pretrained model or train from scratch, please follow the instructions below.
+
+### Download Pretrained Model and Dataset
+
+Download the pretrained checkpoint (`qm9.pt`) and the dataset archive (`qm9_data.tar.gz`) from our [Hugging Face repository](https://huggingface.co/dptech/Uni-3DAR/tree/main).
+
+### Inference with Pretrained Model
+
+To generate QM9 molecules using the pretrained model:
+
+```
+bash inference_qm9.sh qm9.pt
+```
+
+### Train from Scratch
+
+To train the model from scratch:
+
+1. Extract the dataset:
+```
+tar -xzvf qm9_data.tar.gz
+```
+
+2. Run the training script with your desired data path and experiment name:
+
+```
+base_dir=/your_folder_to_save/ bash train_qm9.sh ./qm9_data/ name_of_your_exp
+```
 
 
+Citation
+--------
 
+Please kindly cite our papers if you use the data/code/model.
 
+```
+@article{lu2025uni3dar,
+  author    = {Shuqi Lu and Haowei Lin and Lin Yao and Zhifeng Gao and Xiaohong Ji and Weinan E and Linfeng Zhang and Guolin Ke},
+  title     = {Uni-3DAR: Unified 3D Generation and Understanding via Autoregression on Compressed Spatial Tokens},
+  journal   = {Arxiv},
+  year      = {2025},
+}
+```