Create translation.py

translation.py

@@ -0,0 +1,41 @@
+from transformers import MarianTokenizer, MarianMTModel
+
+# Dictionary of supported languages with MarianMT codes
+LANGUAGES = {
+    "English": "en",
+    "French": "fr",
+    "Spanish": "es",
+    "German": "de",
+    "Chinese": "zh",
+    "Arabic": "ar",
+    "Russian": "ru",
+    "Hindi": "hi",
+    "Japanese": "ja"
+}
+
+# Cache the model loading for faster performance
+@st.cache_resource
+def load_model(src_lang, tgt_lang):
+    """Load the MarianMT model and tokenizer for a language pair."""
+    model_name = f"Helsinki-NLP/opus-mt-{src_lang}-{tgt_lang}"
+    try:
+        tokenizer = MarianTokenizer.from_pretrained(model_name)
+        model = MarianMTModel.from_pretrained(model_name)
+        return tokenizer, model
+    except Exception as e:
+        raise Exception(f"Model for {src_lang} to {tgt_lang} not available: {str(e)}")
+
+def translate(text, source_lang, target_lang):
+    """Translate text from source language to target language."""
+    if not text:
+        return "Please provide text to translate."
+    
+    # Get language codes from dictionary
+    src_code = LANGUAGES.get(source_lang)
+    tgt_code = LANGUAGES.get(target_lang)
+    
+    # Load model and translate
+    tokenizer, model = load_model(src_code, tgt_code)
+    inputs = tokenizer(text, return_tensors="pt", padding=True, truncation=True, max_length=400)
+    translated = model.generate(**inputs)
+    return tokenizer.decode(translated[0], skip_special_tokens=True)