Improve model card and add pipeline tag

Hi, I'm Niels from the Hugging Face community science team. I noticed this model repository could benefit from improved documentation and metadata.

This PR adds:
- The `text-to-video` pipeline tag to the YAML metadata to improve discoverability.
- A more structured README including links to the project page, paper, and official GitHub repository.
- Basic usage instructions (environment setup and inference) adapted from the official README.
- The BibTeX citation for the paper.

These changes help users understand how to run and cite the model effectively!

README.md

@@ -1,9 +1,52 @@
+---
+pipeline_tag: text-to-video
+---
+
 # ShotStream: Streaming Multi-Shot Video Generation for Interactive Storytelling
 
-TL;DR: We propose ShotStream, a novel causal multi-shot architecture that enables interactive storytelling and efficient on-the-fly frame generation, achieving 16 FPS on a single NVIDIA GPU.
+**ShotStream** is a novel causal multi-shot architecture that enables interactive storytelling and efficient on-the-fly frame generation. It achieves sub-second latency and 16 FPS on a single NVIDIA GPU by reformulating the task as next-shot generation conditioned on historical context.
+
+[**Project Page**](https://luo0207.github.io/ShotStream/) | [**Paper**](https://arxiv.org/abs/2603.25746) | [**Code**](https://github.com/KlingAIResearch/ShotStream)
+
+## Introduction
+Multi-shot video generation is crucial for long narrative storytelling. ShotStream allows users to dynamically instruct ongoing narratives via streaming prompts. It preserves visual coherence through a dual-cache memory mechanism and mitigates error accumulation using a two-stage distillation strategy (Distribution Matching Distillation).
+
+## Usage
+
+For the full implementation and training details, please refer to the [official GitHub repository](https://github.com/KlingAIResearch/ShotStream).
+
+### 1. Environment Setup
+
+```bash
+git clone https://github.com/KlingAIResearch/ShotStream.git
+cd ShotStream
+# Setup environment using the provided script
+bash tools/setup/env.sh
+```
+
+### 2. Download Checkpoints
+
+```bash
+# Download the checkpoints of Wan-T2V-1.3B and ShotStream
+bash tools/setup/download_ckpt.sh
+```
+
+### 3. Run Inference
+
+To perform autoregressive 4-step long multi-shot video generation:
+
+```bash
+bash tools/inference/causal_fewsteps.sh
+```
 
-Please refer to the [Github](https://github.com/KlingAIResearch/ShotStream/blob/main/README.md) README for usage.
+## Citation
+If you find our work helpful, please cite our paper:
 
-* Paper：[https://arxiv.org/abs/2603.25746](https://arxiv.org/abs/2603.25746)
-* Project Page：[https://luo0207.github.io/ShotStream/](https://luo0207.github.io/ShotStream/)
-* Code：[https://github.com/KlingAIResearch/ShotStream](https://github.com/KlingAIResearch/ShotStream)
+```bibtex
+@article{luo2026shotstream,
+  title={ShotStream: Streaming Multi-Shot Video Generation for Interactive Storytelling},
+  author={Luo, Yawen and Shi, Xiaoyu and Zhuang, Junhao and Chen, Yutian and Liu, Quande and Wang, Xintao and Pengfei Wan and Xue, Tianfan},
+  journal={arXiv preprint arXiv:2603.25746},
+  year={2026}
+}
+```