app.py

import gradio as gr

def greet(name):
    return "Hello " + name + "!!"

iface = gr.Interface(fn=greet, inputs="text", outputs="text")
# Import the required libraries
from gradio import Interface, gradio as gr

# Define the translation function
def translate_text_to_text(input_text, source_language, target_language, text_to_speech):
    # Initialize the Gradio client
    client = gr.Client("https://facebook-seamless-m4t.hf.space/")
    
    # Define the input parameters for text-to-text translation
    task = "T2TT (Text to Text translation)"
    audio_source = "text"
    input_speech = ""
    
    # Make a prediction for text-to-text translation using the Gradio client
    translation_result = client.predict(
        task,
        audio_source,
        input_speech,
        input_text,
        source_language,
        target_language,
        api_name="/run"
    )
    
    # Extract the translated text from the result
    translated_text = translation_result["output_text"]
    
    # If text-to-speech is enabled, perform text-to-speech synthesis
    if text_to_speech:
        # Define the input parameters for text-to-speech synthesis
        task = "T2ST (Text to Speech translation)"
        audio_source = "text"
        input_speech = translated_text
        
        # Make a prediction for text-to-speech synthesis using the Gradio client
        text_to_speech_result = client.predict(
            task,
            audio_source,
            input_speech,
            "",
            target_language,  # Use the target language for speech synthesis
            target_language,  # Use the target language for speech synthesis
            api_name="/run"
        )
        
        # Extract and return the synthesized speech audio
        synthesized_speech_audio = text_to_speech_result["output_audio"]
        return translated_text, synthesized_speech_audio
    
    # If text-to-speech is not enabled, return only the translated text
    return translated_text

# Create a Gradio interface
iface = Interface(
    fn=translate_text_to_text,
    inputs=[
        gr.Textbox(label="Input Text"),
        gr.Dropdown(label="Source Language", choices=["English", "Spanish", "French", "German"], default="English"),
        gr.Dropdown(label="Target Language", choices=["English", "Spanish", "French", "German"], default="Spanish"),
        gr.Checkbox(label="Enable Text-to-Speech", default=False)
    ],
    outputs=[
        gr.Textbox(label="Translated Text"),
        gr.Audio(label="Synthesized Speech", type="numpy", src="output[1]")
    ]  # Use src="output[1]" to display the synthesized speech
)

# Create a Gradio interface with a custom description
custom_description = """
<div style="text-align:center;">
    <h1>Text Translation and Text-to-Speech Application</h1>
    <h2>Enabled with English, Spanish, French, German, Urdu and Hindi</h2>
    <p>Created by Syed Usman Ghani :: NED University of Engineering & Technology :: Python for Data Science Course :: Mid-term Project</p>
</div>
"""

iface = Interface(
    fn=translate_text_to_text,
    inputs=[
        gr.Textbox(label="Input Text"),
         gr.Dropdown(label="Source Language", choices=["English", "Spanish", "French", "German", "Urdu", "Hindi"], default="English"),
        gr.Dropdown(label="Target Language", choices=["English", "Spanish", "French", "German", "Urdu", "Hindi"], default="Spanish"),
        gr.Checkbox(label="Enable Text-to-Speech", default=False)
    ],
    outputs=[
        gr.Textbox(label="Translated Text"),
        gr.Audio(label="Synthesized Speech", type="numpy", src="output[1]")
    ],
    title="Text & Speech Translation App",  # Set a title for your app
    description=custom_description,  # Use the custom description HTML
)


# Launch the Gradio interface
iface.launch(share=True)