Update app.py

app.py

@@ -19,27 +19,6 @@ model.eval()  # Set the model to evaluation mode
 # Initialize the inference client (if needed for other API-based tasks)
 client = InferenceClient(token=access_token)
 
-
-# Import required modules for E2-F5-TTS
-from huggingface_hub import Client
-
-# Initialize the E2-F5-TTS client
-client_tts = Client("mrfakename/E2-F5-TTS")
-
-def text_to_speech(text, sample):
-    result = client_tts.predict(
-        ref_audio_input=handle_file(f'input/{sample}.mp3'),
-        ref_text_input="",
-        gen_text_input=text,
-        remove_silence=False,
-        cross_fade_duration_slider=0.15,
-        speed_slider=1,
-        api_name="/basic_tts"
-    )
-    audio_file = open(result[0], "rb")
-    audio_bytes = audio_file.read()
-    return audio_bytes
-
 def conversation_predict(input_text):
     """Generate a response for single-turn input using the model."""
     # Tokenize the input text
@@ -49,12 +28,7 @@ def conversation_predict(input_text):
     outputs = model.generate(input_ids, max_new_tokens=2048)
 
     # Decode and return the generated response
-    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
-
-    # Convert the text response to speech using E2-F5-TTS
-    audio_bytes = text_to_speech(response, sample="input")
-    
-    return response, audio_bytes
+    return tokenizer.decode(outputs[0], skip_special_tokens=True)
 
 def respond(
     message: str,