Update README.md

README.md

@@ -111,4 +111,16 @@ Predicted Tokens
 * **Optimizer:** AdamW or Lion (4-bit / mixed precision)
 * **Dataset:** TinyChat (~50–200 MB)
 * **Training loop:** gradient checkpointing + recomputation
-* **Objective:** token prediction / text generation
+* **Objective:** token prediction / text generation
+
+
+## Citation
+
+```bibtex
+@software{lorpt2025,
+  title={LoRPt: Low-Rank Pretraining for Resource-Efficient Language Models},
+  author={[FlameF0X]},
+  year={2025},
+  url={https://github.com/FlameF0X/Low-Rank-Pretraining}
+}
+```