Update README.md

README.md

@@ -1,32 +1,35 @@
----
-pipeline_tag: text-generation
-inference: true
-license: apache-2.0
-datasets:
-- simplescaling/s1K
----
-
-# Model Summary
-
-> s1-0.5B is a reasoning model finetuned from Qwen2.5-0.5B-Instruct on just 1,000 examples. This model was created simply to test the process used to train the original S1 cited below using consumer grade GPUs.
-
-- **Repository:** [simplescaling/s1](https://github.com/simplescaling/s1)
-- **Paper:** https://arxiv.org/abs/2501.19393
-
-# Use
-
-The model usage is documented [here](https://github.com/simplescaling/s1?tab=readme-ov-file#inference).
-
-# Citation
-
-```bibtex
-@misc{muennighoff2025s1simpletesttimescaling,
-      title={s1: Simple test-time scaling}, 
-      author={Niklas Muennighoff and Zitong Yang and Weijia Shi and Xiang Lisa Li and Li Fei-Fei and Hannaneh Hajishirzi and Luke Zettlemoyer and Percy Liang and Emmanuel Candès and Tatsunori Hashimoto},
-      year={2025},
-      eprint={2501.19393},
-      archivePrefix={arXiv},
-      primaryClass={cs.CL},
-      url={https://arxiv.org/abs/2501.19393}, 
-}
-```
+---
+pipeline_tag: text-generation
+inference: true
+license: apache-2.0
+datasets:
+- simplescaling/s1K
+base_model:
+- Qwen/Qwen2.5-0.5B-Instruct
+library_name: transformers
+---
+
+# Model Summary
+
+> s1-0.5B is a reasoning model finetuned from Qwen2.5-0.5B-Instruct on just 1,000 examples. This model was created simply to test the process used to train the original S1 cited below using consumer grade GPUs.
+
+- **Repository:** [simplescaling/s1](https://github.com/simplescaling/s1)
+- **Paper:** https://arxiv.org/abs/2501.19393
+
+# Use
+
+The model usage is documented [here](https://github.com/simplescaling/s1?tab=readme-ov-file#inference).
+
+# Citation
+
+```bibtex
+@misc{muennighoff2025s1simpletesttimescaling,
+      title={s1: Simple test-time scaling}, 
+      author={Niklas Muennighoff and Zitong Yang and Weijia Shi and Xiang Lisa Li and Li Fei-Fei and Hannaneh Hajishirzi and Luke Zettlemoyer and Percy Liang and Emmanuel Candès and Tatsunori Hashimoto},
+      year={2025},
+      eprint={2501.19393},
+      archivePrefix={arXiv},
+      primaryClass={cs.CL},
+      url={https://arxiv.org/abs/2501.19393}, 
+}
+```