Create README.md

README.md

@@ -0,0 +1,38 @@
+---
+license: mit
+tags:
+- biology
+---
+
+<div align="center">
+  <!-- TODO: Uncomment and set YOUR_IMAGE_URL -->
+  <!-- <img src="YOUR_IMAGE_URL" width="100%" alt="OneGenome-Rice (OGR)" /> -->
+  *(Banner / architecture figure: add URL, then uncomment the line above.)*
+</div>
+
+# OneGenome-Rice (OGR)
+
+OGR is a foundational model for AI-driven precision breeding and functional genomics in rice. It is a generative genomic foundation model trained to process DNA sequences up to **1 million** base pairs in length, with **1.25B** total parameters and a **Mixture-of-Experts (MoE)** architecture. It was pre-trained on a curated corpus of **422** rice genomes spanning cultivated and wild *Oryza* diversity.
+
+For instructions, details, and examples, see the project repository: *[TODO: GitHub or documentation URL](https://github.com/TODO/TODO)*.
+
+The table below summarizes training scale and key hyperparameters. **Trained Tokens** follows the **Training Process** section (sequence curriculum and CPT).
+
+<!-- If you ship multiple sizes (e.g. Small / Large), duplicate the table and add columns. -->
+
+| Model Specification | OneGenome-Rice (OGR) |
+| --- | --- |
+| **Model Scale** | |
+| Total Parameters | 1.25B |
+| Activated Parameters | 0.33B |
+| Trained Tokens | ~490B (sequence curriculum) + ~104B (CPT) |
+| **Architecture** | |
+| Architecture | MoE |
+| Number of Experts | 8 |
+| Selected Experts per Token | 2 |
+| Number of Layers | 12 |
+| Attention Hidden Dimension | 1024 |
+| Number of Attention Heads | 16 (GQA, 8 KV groups) |
+| MoE Hidden Dimension (per Expert) | 4096 |
+| Vocabulary Size | 128 (padded) |
+| Context Length | up to 1M |