Add library_name and pipeline_tag to metadata

README.md

@@ -1,6 +1,11 @@
 ---
-license: apache-2.0
 base_model: Qwen/Qwen2.5-VL-7B-Instruct
+datasets:
+- ClareNie/Persona-MME
+- ClareNie/PersonaVLM-Dataset
+license: apache-2.0
+library_name: transformers
+pipeline_tag: image-text-to-text
 tags:
 - multimodal
 - personalized-mllm
@@ -9,9 +14,6 @@ tags:
 - cvpr2026
 - personality-evolving
 - benchmark
-datasets:
-- ClareNie/Persona-MME
-- ClareNie/PersonaVLM-Dataset
 ---
 
 # PersonaVLM: Long-Term Personalized Multimodal LLMs (CVPR 2026)
@@ -45,6 +47,17 @@ This project consists of several components. You can access the model weights, t
 | 📊 **Benchmark** | [HF: Persona-MME (2,000+ cases)](https://huggingface.co/datasets/ClareNie/Persona-MME) |
 | 📂 **Training Data** | [HF: PersonaVLM-Dataset (80k+ samples)](https://huggingface.co/datasets/ClareNie/PersonaVLM-Dataset) |
 
+## 🚀 Quick Start
+Since PersonaVLM is built upon Qwen2.5-VL, it is strictly required to install `transformers==4.51.3` to prevent compatibility issues.
+
+To run inference via CLI:
+```bash
+python inference.py 
+# Optional arguments:
+# --force-retrieve   : Force the agent to execute memory retrieval for every message.
+# --reasoning-mode   : Flag to toggle reasoning (default is True; passing this sets it to False).
+```
+
 ## ✒️ Citation
 If you find our work helpful, please cite our paper:
 
@@ -55,4 +68,5 @@ If you find our work helpful, please cite our paper:
   booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
   year={2026},
   url={http://arxiv.org/abs/2604.13074}
-}
+}
+```