Added talking head functionality back in

app.py

@@ -51,7 +51,7 @@ BUG_FOUND_MSG = "Congratulations, you've found a bug in this application!"
 AUTH_ERR_MSG = "Please paste your OpenAI key from openai.com to use this application. "
 MAX_TOKENS = 512
 
-LOOPING_TALKING_HEAD = "videos/Masahiro.mp4"
+LOOPING_TALKING_HEAD = "videos/Marc.mp4"
 TALKING_HEAD_WIDTH = "192"
 MAX_TALKING_HEAD_TEXT_LENGTH = 155
 
@@ -502,7 +502,7 @@ def do_html_video_speak(words_to_speak, azure_language):
 
     headers = {"Authorization": f"Bearer {os.environ['EXHUMAN_API_KEY']}"}
     body = {
-        'bot_name': 'Masahiro',
+        'bot_name': 'Marc',
         'bot_response': words_to_speak,
         'azure_voice': azure_voice,
         'azure_style': 'friendly',
@@ -521,7 +521,7 @@ def do_html_video_speak(words_to_speak, azure_language):
             f.write(response_stream.read())
         temp_file = gr.File("videos/tempfile.mp4")
         temp_file_url = "/file=" + temp_file.value['name']
-        html_video = f'<video width={TALKING_HEAD_WIDTH} height={TALKING_HEAD_WIDTH} autoplay><source src={temp_file_url} type="video/mp4" poster="Masahiro.png"></video>'
+        html_video = f'<video width={TALKING_HEAD_WIDTH} height={TALKING_HEAD_WIDTH} autoplay><source src={temp_file_url} type="video/mp4" poster="Marc.png"></video>'
     else:
         print('video url unknown')
     return html_video, "videos/tempfile.mp4"
@@ -613,6 +613,23 @@ with gr.Blocks(css="css/custom_css.css") as block:
                                                 show_label=False, lines=1, type='password', elem_id="gr-component")
 
         with gr.Row():
+            with gr.Column(scale=1, min_width=TALKING_HEAD_WIDTH, visible=True):
+                speak_text_cb = gr.Checkbox(label="Enable speech", value=False)
+                speak_text_cb.change(update_foo, inputs=[speak_text_cb, speak_text_state],
+                                     outputs=[speak_text_state])
+
+                my_file = gr.File(label="Upload a file", type="file", visible=False)
+                tmp_file = gr.File(LOOPING_TALKING_HEAD, visible=False)
+                # tmp_file_url = "/file=" + tmp_file.value['name']
+                htm_video = create_html_video(LOOPING_TALKING_HEAD, TALKING_HEAD_WIDTH)
+                video_html = gr.HTML(htm_video)
+
+                # my_aud_file = gr.File(label="Audio file", type="file", visible=True)
+                tmp_aud_file = gr.File("audios/tempfile.mp3", visible=False)
+                tmp_aud_file_url = "/file=" + tmp_aud_file.value['name']
+                htm_audio = f'<audio><source src={tmp_aud_file_url} type="audio/mp3"></audio>'
+                audio_html = gr.HTML(htm_audio)
+            
             with gr.Column(scale=7):
                 chatbot = gr.Chatbot(elem_id="gr-component")
 
@@ -634,13 +651,13 @@ with gr.Blocks(css="css/custom_css.css") as block:
         trace_chain_cb.change(update_foo, inputs=[trace_chain_cb, trace_chain_state],
                               outputs=[trace_chain_state])
 
-        # speak_text_cb = gr.Checkbox(label="Speak text from agent", value=False)
-        # speak_text_cb.change(update_foo, inputs=[speak_text_cb, speak_text_state],
-        #                      outputs=[speak_text_state])
+        speak_text_cb = gr.Checkbox(label="Speak text from agent", value=False)
+        speak_text_cb.change(update_foo, inputs=[speak_text_cb, speak_text_state],
+                             outputs=[speak_text_state])
 
-        # talking_head_cb = gr.Checkbox(label="Show talking head", value=True)
-        # talking_head_cb.change(update_talking_head, inputs=[talking_head_cb, talking_head_state],
-        #                       outputs=[talking_head_state, video_html])
+        talking_head_cb = gr.Checkbox(label="Show talking head", value=True)
+        talking_head_cb.change(update_talking_head, inputs=[talking_head_cb, talking_head_state],
+                               outputs=[talking_head_state, video_html])
 
         # monologue_cb = gr.Checkbox(label="Babel fish mode (translate/restate what you enter, no conversational agent)",
         #                           value=False)
@@ -657,8 +674,8 @@ with gr.Blocks(css="css/custom_css.css") as block:
                                  surprise_level_state, sadness_level_state, disgust_level_state, anger_level_state,
                                  lang_level_state, translate_to_state, literary_style_state,
                                  qa_chain_state, docsearch_state, use_embeddings_state],
-                   # outputs=[chatbot, history_state, video_html, my_file, audio_html, tmp_aud_file, message])
-                   outputs=[chatbot, history_state, message])
+                   outputs=[chatbot, history_state, video_html, my_file, audio_html, tmp_aud_file, message])
+                   # outputs=[chatbot, history_state, message])
     # outputs=[chatbot, history_state, audio_html, tmp_aud_file, message])
 
     submit.click(chat, inputs=[openai_api_key_textbox, message, history_state, chain_state, trace_chain_state,
@@ -668,8 +685,8 @@ with gr.Blocks(css="css/custom_css.css") as block:
                                surprise_level_state, sadness_level_state, disgust_level_state, anger_level_state,
                                lang_level_state, translate_to_state, literary_style_state,
                                qa_chain_state, docsearch_state, use_embeddings_state],
-                 # outputs=[chatbot, history_state, video_html, my_file, audio_html, tmp_aud_file, message])
-                 outputs=[chatbot, history_state, message])
+                 outputs=[chatbot, history_state, video_html, my_file, audio_html, tmp_aud_file, message])
+                 # outputs=[chatbot, history_state, message])
     # outputs=[chatbot, history_state, audio_html, tmp_aud_file, message])
 
     openai_api_key_textbox.change(set_openai_api_key,