Create app.py

app.py

@@ -0,0 +1,81 @@
+import whisper
+from TTS.api import TTS
+import requests
+import gradio as gr
+from pydub import AudioSegment
+from deep_translator import GoogleTranslator
+
+# Initialize models
+whisper_model = whisper.load_model("small")  # Faster Whisper model
+tts_model = TTS(model_name="tts_models/en/ljspeech/tacotron2-DDC", progress_bar=False, gpu=False)
+
+# Groq API Key and Base URL
+groq_api_key = "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"
+groq_base_url = "https://api.groq.com"  # Base URL from Groq documentation
+
+# Functions for the Chatbot
+def voice_to_text(audio_path):
+    """Convert voice input to text and detect language using Whisper."""
+    result = whisper_model.transcribe(audio_path)
+    detected_language = result["language"]
+    return result["text"], detected_language
+
+def process_text_with_groq(input_text):
+    """Process user text input using Groq LLM."""
+    url = f"{groq_base_url}/chat/completions"  # Update endpoint if needed
+    headers = {"Authorization": f"Bearer {groq_api_key}"}
+    payload = {
+        "messages": [{"role": "user", "content": input_text}],
+        "model": "llama3-8b-8192",
+        "stream": False,
+    }
+    try:
+        response = requests.post(url, json=payload, headers=headers)
+        response.raise_for_status()  # Raise an error for HTTP issues
+        return response.json()["choices"][0]["message"]["content"]
+    except requests.exceptions.RequestException as e:
+        return f"Error: {str(e)}"
+
+def translate_text(text, target_lang):
+    """Translate text to the target language using deep-translator."""
+    translated_text = GoogleTranslator(source="auto", target=target_lang).translate(text)
+    return translated_text
+
+def text_to_voice(output_text, language_code):
+    """Convert text response to voice using Coqui TTS."""
+    audio_path = "response.wav"
+    tts_model.tts_to_file(text=output_text, file_path=audio_path)
+    return audio_path
+
+# Gradio Interface
+def chatbot(audio_file):
+    # Step 1: Convert audio to text and detect language
+    user_input, detected_language = voice_to_text(audio_file)
+
+    # Step 2: Process the text with Groq LLM
+    bot_response = process_text_with_groq(user_input)
+
+    # Step 3: Translate the response if the detected language is not English
+    if detected_language != "en":  # Translate only if language is not English
+        bot_response = translate_text(bot_response, detected_language)
+
+    # Step 4: Convert the response to voice
+    audio_response_path = text_to_voice(bot_response, detected_language)
+
+    return bot_response, audio_response_path
+
+# Gradio UI
+ui = gr.Interface(
+    fn=chatbot,
+    inputs=gr.Audio(type="filepath", label="Upload Audio File"),
+    outputs=[
+        gr.Textbox(label="Chatbot Response"),
+        gr.Audio(label="Chatbot Voice Response")
+    ],
+    title="Zeeshan Voice-to-Voice Chatbot",
+    description="Upload an audio file to interact with Zeeshan. Zeeshan will listen, process your query, and respond in the same language with both text and voice."
+)
+
+# Launch Gradio app
+if __name__ == "__main__":
+    ui.launch()