Update app.py

app.py

@@ -588,59 +588,30 @@ latent_map = {}
 latent_map["Julian"] = get_latents("voices/julian-bedtime-style-1.wav")
 latent_map["Pirate"] = get_latents("voices/pirate_by_coqui.wav")
 
-#### GRADIO INTERFACE ####
 
-with gr.Blocks(title=title) as demo:
-    chatbot = gr.Chatbot(
-        [],
-        elem_id="chatbot",
-        bubble_full_width=False,
-    )
-
-    chatbot_role = gr.Dropdown(
-        label="Role of the Chatbot",
-        info="How should Chatbot talk like",
-        choices=ROLES,
-        max_choices=1,
-        value=ROLES[0],
-    )
-
-    txt = gr.Textbox(
-        scale=3,
-        show_label=False,
-        placeholder="Enter text and press enter, or speak to your microphone",
-        container=False,
-        interactive=True,
-    )
-    txt_btn = gr.Button(value="Submit text", scale=1)
-
-    with gr.Row():
-        sentence = gr.Textbox(visible=False)
-        audio = gr.Audio(
-            value=None,
-            label="Generated audio response",
-            streaming=True,
-            autoplay=True,
-            interactive=False,
-            show_label=True,
-        )
-
-    def clear_inputs(chatbot):
-        return None
-    clear_btn = gr.ClearButton([chatbot, audio])
-    chatbot_role.change(fn=clear_inputs, inputs=[chatbot], outputs=[chatbot])
-
-    txt_msg = txt_btn.click(add_text, [chatbot, txt], [chatbot, txt], queue=False).then(
-        generate_speech,  [chatbot,chatbot_role], [chatbot,chatbot_role, sentence, audio]
-    )
-
-    txt_msg.then(lambda: gr.update(interactive=True), None, [txt], queue=False)
-
-    txt_msg = txt.submit(add_text, [chatbot, txt], [chatbot, txt], queue=False).then(
-        generate_speech,  [chatbot,chatbot_role], [chatbot,chatbot_role, sentence, audio]
-    )
-
-    txt_msg.then(lambda: gr.update(interactive=True), None, [txt], queue=False)
+# Define the main function for the API endpoint that takes the input text and chatbot role
+def generate_story_and_speech(input_text, chatbot_role):
+    # We assume that other necessary components have been initialized and are ready to use here
+    
+    # Here, we'll integrate the story generation, language detection, and speech synthesis logic
+    # Let's assume `generate_story()` is a function that generates the story based on the input text
+    # And `synthesize_speech()` is a function that synthesizes speech from text
+    story_text = generate_story(input_text, chatbot_role)
+    language = detect_language(story_text)
+    speech_audio_bytes = synthesize_speech(story_text, language)
+
+    # Convert the speech to base64 to include in the JSON response
+    speech_audio_base64 = base64.b64encode(speech_audio_bytes).decode('utf8')
+
+    # Return the story and speech audio in base64 format
+    return {"text": story_text, "audio": speech_audio_base64}
+
+# Create a Gradio Interface using only the `generate_story_and_speech()` function and the 'json' output type
+demo = gr.Interface(
+    fn=generate_story_and_speech,
+    inputs=[gr.Textbox(placeholder="Enter your text here"), gr.Dropdown(choices=ROLES, label="Select Chatbot Role")],
+    outputs="json"
+)
 
 demo.queue()
 demo.launch(debug=True)