Update app.py

app.py

@@ -7,6 +7,7 @@ import gradio as gr
 from collections import defaultdict
 from uuid import uuid4
 
+
 DB_CONN = pymongo.MongoClient(os.environ.get("DB_URL"))
 line_details = {l["source_id"]:l for l in list(DB_CONN['vo_data_dump']['qa_video_dump'].find()) if l.get("done") is not True}
 
@@ -37,13 +38,10 @@ def render_line_details(line_id):
     translation = line_detail["translator_translated_text"]
     response = requests.post(os.environ["CF_URL"], params={"url": line_detail["source_audio_url"]})
     signed_uri = response.text
-    audio_response = requests.get(signed_uri)
-    # path = os.path.join("temp_audios", f"{uuid4()}.mp3")
-    # with open(path, "wb") as f:
-    #     f.write(audio_response.content)
-    return audio_response.content, source_transcript, translation
+    audio_response = requests.get(signed_uri).content
+    return audio_response, source_transcript, translation
 
-def dump_data_db(line_id, video_type, video_subject, video_topics, video_tone, gender, demography, total_time, source_language, target_language, comments, accent, source_transcript, translation):
+def dump_data_db(line_id, video_type, video_subject, video_topics, video_tone, gender, demography, total_time, source_language, target_language, comments, accent, source_transcript, translation, speaker_gender):
     line_detail = line_details[line_id]
     video_id = line_detail["video_id"]
     line_data = {
@@ -55,6 +53,7 @@ def dump_data_db(line_id, video_type, video_subject, video_topics, video_tone, g
         "accent": accent,
         "speaking_rate": round(len(translation.split()) / line_detail["line_duration"], 3),
         "src_time": line_detail["line_duration"],
+        "gender": speaker_gender,
         "version": "v1",
         "done": True
     }
@@ -105,8 +104,8 @@ with gr.Blocks() as demo:
         video_html = gr.HTML(label="Video")
         video_playable = gr.Video(label="Video")
     with gr.Row():
-        video_type_dropdown = gr.Dropdown(["learning", "audiobook", "podcast", "vlog", "news", "cooking", "review"], label="Video Type", value="learning")
-        video_subject_dropdown = gr.Dropdown(["tech", "science", "lifestyle", "cooking", "travel", "finance", "politics"], label="Video Subject", value="tech")
+        video_type_dropdown = gr.Dropdown(["learning", "audiobook", "podcast", "vlog", "news", "cooking", "review"], label="Video Type", value="learning", allow_custom_value=True)
+        video_subject_dropdown = gr.Dropdown(["tech", "science", "lifestyle", "cooking", "travel", "finance", "politics"], label="Video Subject", value="tech", allow_custom_value=True)
         video_topics_text = gr.Textbox(label="Video Topics")
     with gr.Row():
         video_tone_dropdown = gr.Dropdown(["casual", "semi-formal", "formal"], label="Video Tone", value="semi-formal")
@@ -123,6 +122,7 @@ with gr.Blocks() as demo:
     with gr.Row():
         audio_item = gr.Audio(label="Source Audio", type="filepath")
         accent_text = gr.Textbox(label="Accent", value="indian")
+        speaker_gender_dropdown = gr.Dropdown(["female", "male"], label="Speaker Gender", value="female")
     with gr.Row():
         source_transcript_text = gr.Textbox(label="Source Transcript")
         translation_text = gr.Textbox(label="Translation", interactive=False)
@@ -131,8 +131,8 @@ with gr.Blocks() as demo:
     # Actions
     video_id_dropdown.change(render_video_details, video_id_dropdown, [video_title_text, source_language_text, target_language_text, total_time_text, line_id_dropdown, video_html, video_playable])
     line_id_dropdown.change(render_line_details, line_id_dropdown, [audio_item, source_transcript_text, translation_text])
-    update_button.click(dump_data_db, [line_id_dropdown, video_type_dropdown, video_subject_dropdown, video_topics_text, video_tone_dropdown, gender_dropdown, demography_text, total_time_text, source_language_text, target_language_text, comments_text, accent_text, source_transcript_text, translation_text], [line_id_dropdown, video_id_dropdown])
+    update_button.click(dump_data_db, [line_id_dropdown, video_type_dropdown, video_subject_dropdown, video_topics_text, video_tone_dropdown, gender_dropdown, demography_text, total_time_text, source_language_text, target_language_text, comments_text, accent_text, source_transcript_text, translation_text, speaker_gender_dropdown], [line_id_dropdown, video_id_dropdown])
 
 
 if __name__=="__main__":
-    demo.queue().launch(auth=(os.environ.get("USERNAME"), os.environ.get("PASSWORD")))
+    demo.queue().launch(share=True, auth=(os.environ.get("USERNAME"), os.environ.get("PASSWORD")))