app.py

import streamlit as st
from transformers import MarianMTModel, MarianTokenizer

# Load the MarianMT model and tokenizer
model_name = "Helsinki-NLP/opus-mt-en-ROMANCE"
model = MarianMTModel.from_pretrained(model_name)
tokenizer = MarianTokenizer.from_pretrained(model_name)

# Define target language codes directly
language_codes = {
    'French': 'fr',
    'German': 'de',
    'Italian': 'it',
    'Portuguese': 'pt',
    'Spanish': 'es',
    # Add more languages if needed
}

def translate_text(text, target_lang_code):
    # Prepare the input and translate
    inputs = tokenizer.encode(text, return_tensors="pt")
    # Set the decoder start token ID based on language code
    decoder_start_token_id = tokenizer.convert_tokens_to_ids(f"<{target_lang_code}>")
    translated = model.generate(inputs, decoder_start_token_id=decoder_start_token_id)
    translated_text = tokenizer.decode(translated[0], skip_special_tokens=True)
    return translated_text

def main():
    st.title("English to Any Language Translator")

    # User input for English text
    text_to_translate = st.text_area("Enter text in English:")

    # Language options
    selected_language = st.selectbox("Select target language:", list(language_codes.keys()))

    if st.button("Translate"):
        if text_to_translate:
            target_lang_code = language_codes[selected_language]
            translated_text = translate_text(text_to_translate, target_lang_code)
            st.write(f"**Translation in {selected_language}:**")
            st.write(translated_text)
        else:
            st.warning("Please enter text to translate.")

if __name__ == "__main__":
    main()