Evaluation on all 26 language pairs with BLEU, speed, radar

.gitattributes

@@ -34,3 +34,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
 tokenizer.json filter=lfs diff=lfs merge=lfs -text
+images/radar.png filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,73 @@
+---
+language:
+- multilingual
+license: apache-2.0
+library_name: transformers
+tags:
+- translation
+- mt5
+- multilingual
+- whirlwindai
+pipeline_tag: translation
+---
+
+# 🌪️ WhirlwindAI/Translate-25L
+
+A multilingual translation model fine-tuned from `google/mt5-small`, covering 25+ language pairs sourced from OPUS-100.
+
+## 🚀 Usage
+
+```python
+from transformers import MT5ForConditionalGeneration, AutoTokenizer
+
+model = MT5ForConditionalGeneration.from_pretrained("WhirlwindAI/Translate-25L")
+tokenizer = AutoTokenizer.from_pretrained("WhirlwindAI/Translate-25L")
+
+def translate(text, src, tgt):
+    prompt = f"translate {src} to {tgt}: {text}"
+    inputs = tokenizer(prompt, return_tensors="pt", truncation=True)
+    outputs = model.generate(**inputs, max_new_tokens=128, num_beams=4)
+    return tokenizer.decode(outputs[0], skip_special_tokens=True)
+
+print(translate("Hello, how are you?", "en", "fr"))
+```
+
+## 📊 Evaluation (on OPUS-100 test, 50 examples per pair)
+
+### BLEU Scores
+![BLEU Scores](images/bleu_scores.png)
+
+### Inference Speed
+![Speed](images/speed.png)
+
+### Radar Comparison
+![Radar](images/radar.png)
+
+### Sample Translation (French → English)
+| Input (French) | Output (English) |
+|----------------|------------------|
+| `Bonjour, comment allez-vous aujourd'hui ?`  | `Hello, how are you here now?`    |
+
+## 📋 Model Details
+
+| Property | Value |
+|---|---|
+| Base model | google/mt5-small |
+| Parameters | 300M |
+| Training data | OPUS-100 |
+| Languages | 25+ |
+| License | Apache 2.0 |
+
+## 🏆 Highlights
+
+- Up to **41.2** BLEU on en→fr
+- **52.2** tokens/sec average inference speed
+- Lightweight (300M params), fast, and multilingual
+
+## 🙏 Acknowledgments
+
+Built by WhirlwindAI. Fine‑tuned on OPUS-100, powered by Hugging Face.
+
+---
+
+*Translate the world, 25 languages at a time.* 🌍