Update README.md

README.md

@@ -12,15 +12,11 @@ tags:
 - whirlwindai
 ---
 <p align="center">
-
-<img src="https://capsule-render.vercel.app/api?type=blur&height=240&text=Translate-25L&fontSize=55&fontColor=ffffff&animation=twinkling&color=gradient&customColorList=12,20,24,30"/>
-
+<img src="https://capsule-render.vercel.app/api?type=blur&height=240&text=Translate-25T&fontSize=55&fontColor=ffffff&animation=twinkling&color=gradient&customColorList=12,20,24,30"/>
 </p>
 
 <div align="center">
-
-<img src="https://readme-typing-svg.demolab.com?font=Space+Grotesk&weight=700&size=28&duration=2500&pause=1000&color=06B6D4&center=true&vCenter=true&width=850&lines=Translate-25L;Multilingual+Translation.+25%2B+Languages.;Built+for+Speed+and+Clarity." />
-
+<img src="https://readme-typing-svg.demolab.com?font=Space+Grotesk&weight=700&size=28&duration=2500&pause=1000&color=06B6D4&center=true&vCenter=true&width=850&lines=Translate-25T;Multilingual+Translation.+18%2B+Languages.;Built+for+Speed+and+Clarity." />
 <br>
 
 <img src="https://img.shields.io/badge/Parameters-300M-06B6D4?style=for-the-badge">
@@ -34,14 +30,11 @@ tags:
 
 # The Idea
 
-Translate‑25L is a multilingual translation model fine‑tuned from **google/mt5‑small** on the **OPUS‑100** dataset. It supports translation between **English and 25+ other languages** in both directions, making it a practical tool for research, education, and lightweight multilingual applications.
-
+# The Idea
+Translate‑25T is a multilingual translation model fine‑tuned from **google/mt5‑small** on the **OPUS‑100** dataset. It supports translation between **English and 18+ other languages** in both directions, making it a practical tool for research, education, and lightweight multilingual applications.
 The focus is on **balance**: delivering solid translation quality while keeping inference fast and the model compact enough to run in resource‑constrained environments.
-
 <p align="center">
-
 <img src="https://capsule-render.vercel.app/api?type=rect&height=4&color=gradient&customColorList=12,20,24,30"/>
-
 </p>
 
 ---
@@ -72,8 +65,8 @@ print(translate("Hello, how are you?", "en", "fr"))
 
 # Supported Languages
 
-Translate‑25L covers a diverse set of languages, including many high‑resource European languages, as well as several Asian and Middle Eastern languages.
-
+# Supported Languages
+Translate‑25T covers a diverse set of languages, including many high‑resource European languages, as well as several Asian and Middle Eastern languages.
 | Code | Language | Code | Language |
 |------|----------|------|----------|
 | 🇬🇧 en | English | 🇩🇪 de | German |
@@ -85,8 +78,7 @@ Translate‑25L covers a diverse set of languages, including many high‑resourc
 | 🇮🇳 hi | Hindi | 🇮🇩 id | Indonesian |
 | 🇹🇷 tr | Turkish | 🇻🇳 vi | Vietnamese |
 | 🇵🇱 pl | Polish | 🇺🇦 uk | Ukrainian |
-| 🇷🇴 ro | Romanian | 🇨🇿 cs | Czech |
-| 🇸🇪 sv | Swedish | ...and more | |
+| 🇷🇴 ro | Romanian | 🇸🇪 sv | Swedish |
 <p align="center">
 
 <img src="https://capsule-render.vercel.app/api?type=rect&height=4&color=gradient&customColorList=12,20,24,30"/>
@@ -97,7 +89,7 @@ Translate‑25L covers a diverse set of languages, including many high‑resourc
 
 # Evaluation
 
-We evaluated Translate‑25L on the **OPUS‑100 test set** using **50 samples per language pair**. The results below show BLEU scores, inference speed, and a combined radar view.
+We evaluated Translate‑25T on the **OPUS‑100 test set** using **50 samples per language pair**. The results below show BLEU scores, inference speed, and a combined radar view.
 
 ### BLEU Scores
 ![BLEU Scores](images/bleu_scores.png)
@@ -137,7 +129,7 @@ We evaluated Translate‑25L on the **OPUS‑100 test set** using **50 samples p
 | Parameters | 300 Million |
 | Architecture | mT5 (Encoder‑Decoder) |
 | Training Data | OPUS-100 |
-| Languages | 25+ |
+| Languages | 18+ |
 | Framework | Hugging Face Transformers |
 | License | Apache 2.0 |
 <p align="center">
@@ -149,16 +141,13 @@ We evaluated Translate‑25L on the **OPUS‑100 test set** using **50 samples p
 ---
 
 # Highlights
-
-- **Multilingual**: translate between English and 25+ languages.
+- **Multilingual**: translate between English and 18+ languages.
 - **Efficient**: ~300M parameters – lightweight enough for many production scenarios.
 - **Fast**: average inference speed of **52.2 tokens/second** on a T4 GPU.
 - **Research‑friendly**: open weights and Apache 2.0 license.
 - **Practical**: trained on a diverse set of parallel sentences from OPUS‑100.
 <p align="center">
-
 <img src="https://capsule-render.vercel.app/api?type=rect&height=4&color=gradient&customColorList=12,20,24,30"/>
-
 </p>
 
 ---