Update README.md

README.md

@@ -85,3 +85,17 @@ stream = client.chat.completions.create(
 * Steiner’s current post-training data does not include examples for multi-turn dialogues. The best-performing version of the Steiner model (based on [Qwen2.5-32B](https://huggingface.co/Qwen/Qwen2.5-32B)) lacks the ability to handle multi-turn conversations. The open-source Steiner-preview model (based on [Qwen2.5-32B-Instruct](https://huggingface.co/Qwen/Qwen2.5-32B-Instruct)) is compatible with chat formats but is still not recommended for multi-turn dialogues.
 * Similar to OpenAI o1-2024-09-12, Steiner also does not recommend the use of custom system prompts or modifications to sampling parameters such as temperature. Steiner has not yet been trained on a diverse set of system prompts, and altering other parameters may lead to errors in the formatting of reasoning tokens.
 * The language composition of Steiner's post-training data is approximately 90% English and 10% Chinese, but during the reasoning path data augmentation process, almost only English was used. Therefore, while the model's final responses demonstrate a certain degree of language following ability, the reasoning tokens may predominantly be generated in English.
+
+## Citation
+
+If you find my work helpful, please consider citing it in your research or projects. Your acknowledgment would be greatly appreciated!
+
+```
+@misc{ji2024steiner,
+    title = {A Small Step Towards Reproducing OpenAI o1: Progress Report on the Steiner Open Source Models},
+    url = {https://medium.com/@peakji/b9a756a00855},
+    author = {Yichao Ji},
+    month = {October},
+    year = {2024}
+}
+```