Add model card and metadata for AnomaMind-8B

README.md

@@ -0,0 +1,43 @@
+---
+license: mit
+library_name: transformers
+pipeline_tag: other
+base_model: Qwen/Qwen3-8B
+---
+
+# AnomaMind-8B
+
+AnomaMind is an agentic framework that reformulates time series anomaly detection (TSAD) as an evidence-driven sequential decision-making process. This repository contains the 8B detector model, which is based on the **Qwen3-8B** architecture and optimized using reinforcement learning.
+
+The model was introduced in the paper: [AnomaMind: Agentic Time Series Anomaly Detection with Tool-Augmented Reasoning](https://huggingface.co/papers/2602.13807).
+
+## Resources
+- **Paper:** [AnomaMind: Agentic Time Series Anomaly Detection with Tool-Augmented Reasoning](https://huggingface.co/papers/2602.13807)
+- **Repository:** [Xiaoyu-Tao/AnomaMind-TS](https://github.com/Xiaoyu-Tao/AnomaMind-TS)
+
+## Model Description
+
+AnomaMind operates through a coarse-to-fine workflow that localizes suspicious intervals, constructs diagnostic evidence through tool interaction (statistical, value-based, and change-based operators), and refines decisions through self-reflection. This hybrid approach allows general-purpose models to handle flexible reasoning while this task-specific policy handles precise detection.
+
+## Usage
+
+### Serving with vLLM
+
+You can serve the model using [vLLM](https://github.com/vllm-project/vllm) with the following command (replace the path with your local checkpoint if necessary):
+
+```bash
+vllm serve YuChongZ/AnomaMind-8B --port 8000 --max-model-len 11000 --gpu-memory-utilization 0.95 --enable-auto-tool-choice --tool-call-parser hermes --served-model-name detector
+```
+
+Once served, the model can be used for inference as part of the AnomaMind agentic workflow. For detailed environment setup and data preprocessing, please refer to the [official GitHub repository](https://github.com/Xiaoyu-Tao/AnomaMind-TS).
+
+## Citation
+
+```bibtex
+@article{tao2026anomamind,
+  title={AnomaMind: Agentic Time Series Anomaly Detection with Tool-Augmented Reasoning},
+  author={Tao, Xiaoyu and Wu, Yuchong and Cheng, Mingyue and Guo, Ze and Gao, Tian},
+  journal={arXiv preprint arXiv:2602.13807},
+  year={2026}
+}
+```