Update app.py

app.py

@@ -2,6 +2,7 @@ import gradio as gr
 import torch
 from transformers import AutoProcessor, BarkModel
 import scipy
+from pytube import YouTube
 
 
 # device = "cuda" if torch.cuda.is_available() else "cpu"
@@ -34,15 +35,23 @@ def run_bark(text, n, lang):
     scipy.io.wavfile.write("bark_out.wav", rate=sampling_rate, data=speech_values.cpu().numpy().squeeze())
     return ("bark_out.wav")
 
+def load_video_yt(vid):
+    yt = YouTube(vid)
+    vid = yt.streams.filter(progressive=True, file_extension='mp4').order_by('resolution').desc().first().download(filename="tmp.mp4")
+    vid_aud = yt.streams.filter(only_audio=True)[0].download(filename="tmp_aud.mp3")
+    return vid, vid_aud 
+    
 with gr.Blocks() as app:
     with gr.Column():
         in_text = gr.Textbox()
         with gr.Tab("Default"):
-            speaker_num = gr.Dropdown(label="Speaker Voice", choices=num_list,value="1")
-            speaker_lang = gr.Dropdown(label="Speaker Language", choices=lang_list,value="en")
+            with gr.Row():
+                speaker_num = gr.Dropdown(label="Speaker Voice", choices=num_list,value="1")
+                speaker_lang = gr.Dropdown(label="Speaker Language", choices=lang_list,value="en")
         with gr.Tab("Upload"):
-            in_aud_mic = gr.Audio(source='microphone')
-            in_aud_file = gr.Audio(source='upload')
+            with gr.Row():
+                in_aud_mic = gr.Audio(source='microphone')
+                in_aud_file = gr.Audio(source='upload')
             in_aud_yout = gr.Textbox(label="YouTube URL")
             yout_vid = gr.Video()
         #speaker_num = gr.Number(value=0)