update

app.py

@@ -6,7 +6,7 @@ import uuid
 import whisper
 import azure.cognitiveservices.speech as speechsdk
 import base64
-
+import os
 
 from polly_utils import PollyVoiceData, NEURAL_ENGINE
 from azure_utils import AzureVoiceData
@@ -323,11 +323,11 @@ with gr.Blocks(css=css) as demo:
                     max_tokens = gr.Slider(minimum=100, maximum=4096, value=1000, step=1, label="Max tokens per response")
                     context_length = gr.Slider(minimum=1, maximum=10, value=2, step=1, label="Context length", info="Number of previous messages to send to the chatbot. Be careful with high values, it can blow up the token budget quickly.")
         with gr.Row():
-            with gr.Column(scale=2, min_width=0):
+            with gr.Column(min_width=TALKING_HEAD_WIDTH, visible=True):
                 type_select = gr.Dropdown(show_label=False, choices= ["TEXT", "IMAGE"],value="TEXT",interactive=True)
-            with gr.Column(scale=8):
+            with gr.Column(scale=6):
                 input_message = gr.Textbox(show_label=False, placeholder="Enter text and press enter", visible=True).style(container=False)
-            with gr.Column(scale=8):
+            with gr.Column(scale=3):
                 btn_submit = gr.Button("Submit")
                 total_tokens_str = gr.Markdown(elem_id="total_tokens_str")
         with gr.Row():