Update README.md

README.md

@@ -1,35 +1,36 @@
 ---
-title: FormosanBank
-emoji: 🌖
-colorFrom: pink
-colorTo: green
-sdk: static
-pinned: false
-license: other
+title: README  
+emoji: 🌖  
+colorFrom: pink  
+colorTo: green  
+sdk: static  
+pinned: false  
 ---
 
-# FormosanBank
+## FormosanBank
 
-FormosanBank is a centralized repository of corpora and tooling for the 16 extant Formosan (Taiwan Indigenous) languages—designed for research, education, and community use. The documentation covers data structure, access methods, and project background. :contentReference[oaicite:0]{index=0}
+**What is FormosanBank?**  
+FormosanBank is an open-source repository of corpora and quality-control tools supporting the documentation, processing, and machine-learning use of Taiwan’s indigenous Formosan languages. :contentReference[oaicite:1]{index=1}
 
-- **Scale (at a glance):** >8M tokens and ~730 hours of audio across the languages (see GitBook “Welcome” for breakdowns). :contentReference[oaicite:1]{index=1}  
-- **What’s inside:** texts & transcriptions, dictionaries, reference grammars, and curated corpora entries (e.g., NTU Paiwan ASR, ePark, Wikipedias). :contentReference[oaicite:2]{index=2}  
-- **Data format:** a standardized XML schema (inspired by the Pangloss Collection) to ensure consistent metadata and downstream processing. :contentReference[oaicite:3]{index=3}
+**Key Features:**  
+- A large collection of corpora across multiple Formosan languages (e.g., Amis, Paiwan, Atayal) with scripts for cleaning, orthography extraction, validation etc. :contentReference[oaicite:2]{index=2}  
+- Quality control modules: punctuation checks, non-ASCII filtering, XML-template verification, orthography extraction. :contentReference[oaicite:3]{index=3}  
+- Designed to support downstream NLP tasks (translation, ASR, summarization) for low-resource languages. :contentReference[oaicite:4]{index=4}  
+- License: *[you should insert your specific license here]*  
+- Maintained in a GitHub repository: [https://github.com/FormosanBank/FormosanBank](https://github.com/FormosanBank/FormosanBank) :contentReference[oaicite:5]{index=5}  
+- Linked documentation: [https://ai4commsci.gitbook.io/formosanbank](https://ai4commsci.gitbook.io/formosanbank)  
 
-## Documentation
+**Usage / SDK:**  
+Since the card says `sdk: static` this suggests you are using static hosting of docs or a simple web UI. You can embed links to the repo, the docs, usage instructions etc.
 
-- **GitBook (primary docs):** https://ai4commsci.gitbook.io/formosanbank  
-  Start with *Background → Why Formosan*, *Developers*, *Corpora*, and the *XML Format* pages. :contentReference[oaicite:4]{index=4}
+**Short description:**  
+Building open-source infrastructure and corpora for Taiwan’s indigenous Formosan languages, enabling machine translation, ASR and summarization efforts in extremely low-resource settings.
 
-## Repository
-
-- **GitHub (code & corpora structure):** https://github.com/FormosanBank/FormosanBank  
-  Includes quality-control utilities (XML validation, orthography comparison, cleaning scripts) and corpus organization. :contentReference[oaicite:5]{index=5}
-
-## Terms of Use (Summary)
-
-You’re free to use and redistribute FormosanBank materials **with restrictions**, including **No Commercial AI Use** (see full Terms for details). If in doubt, consult the GitBook Terms page. :contentReference[oaicite:6]{index=6}
-
-## Cite / Related Work
+---
 
-If you use FormosanBank in publications or demos, please reference the project documentation and any linked corpora. Recent research and benchmarks drawing on FormosanBank’s preparation pipelines include work on low-resource evaluation for Formosan languages. :contentReference[oaicite:7]{index=7}
+### Getting started  
+1. Clone the repository:  
+```bash
+git clone https://github.com/FormosanBank/FormosanBank
+cd FormosanBank
+pip install -r requirements.txt