Jarvis Model

app.py

@@ -43,11 +43,11 @@ async def generate1(prompt):
     for response in stream:
         output += response.token.text
 
-    communicate = edge_tts.Communicate(output)
-    with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as tmp_file:
-        tmp_path = tmp_file.name
-        await communicate.save(tmp_path)
-    yield tmp_path
+    # communicate = edge_tts.Communicate(output)
+    # with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as tmp_file:
+    #     tmp_path = tmp_file.name
+    #     await communicate.save(tmp_path)
+    yield output
 
 client2 = InferenceClient("meta-llama/Meta-Llama-3-70B-Instruct")
 
@@ -104,13 +104,15 @@ with gr.Blocks(css="style.css") as demo:
     with gr.Row():
         user_input = gr.Textbox(label="Prompt", value="What is Wikipedia")
         input_text = gr.Textbox(label="Input Text", elem_id="important")
+        output_text = gr.Textbox(label="Output")
         # output_audio = gr.Audio(label="JARVIS", type="filepath",
         #                 interactive=False,
         #                 autoplay=True,
         #                 elem_classes="audio")
     with gr.Row():
         translate_btn = gr.Button("Response")
-        translate_btn.click(fn=generate1, inputs=user_input, api_name="translate")
+        translate_btn.click(fn=generate1, inputs=user_input,
+                            outputs=output_text, api_name="translate")
         
 gr.Markdown(MORE)