ok

app.py

@@ -2,6 +2,7 @@ import os
 os.system("pip install git+https://github.com/openai/whisper.git")
 import gradio as gr
 import whisper
+from convert import main
 
 model = whisper.load_model("small")
 
@@ -19,6 +20,9 @@ def inference(audio):
     print(result.text)
     return result.text
 
+def inference2(url):
+    return main(url)
+
 
 title="Whisper"
 
@@ -158,6 +162,13 @@ with block:
         text = gr.Textbox(show_label=False)
         
         btn.click(inference, inputs=[audio], outputs=[text])
+
+        # add another textbox to get url from user
+        url = gr.Textbox(label="URL", show_label=False)
+        btn2 = gr.Button("Transcribe")
+        text2 = gr.Textbox(show_label=False)
+        btn2.click(inference2, inputs=[url], outputs=[text2])
+
  
         gr.HTML('''
         <div class="footer">

convert.py

@@ -0,0 +1,58 @@
+import pytube
+from moviepy.editor import VideoFileClip
+import pywhisper
+import os
+
+def download_video(url):
+    video = pytube.YouTube(url)
+    stream = video.streams.get_by_itag(18)
+    stream.download()
+    return stream.default_filename
+
+def convert_to_mp3(filename):
+    clip = VideoFileClip(filename)
+    clip.audio.write_audiofile(filename[:-4] + ".mp3")
+    clip.close()
+
+def AudiotoText(filename):
+    model = pywhisper.load_model("base")
+    result = model.transcribe(filename)
+    print(result["text"])
+    sonuc = result["text"]
+    return sonuc
+
+def main(url):
+    print('''
+    This tool will convert Youtube videos to mp3 files and then transcribe them to text using Whisper.
+    ''')
+
+    print("Downloading video... Please wait.")
+    try:
+        filename = download_video(url)
+        print("Downloaded video as " + filename)
+    except:
+        print("Not a valid link..")
+        return
+    try:
+        convert_to_mp3(filename)
+        print("Converted video to mp3")
+    except:
+        print("Error converting video to mp3")
+        return
+    try:
+        model = pywhisper.load_model("base")
+        result = model.transcribe(filename[:-4] + ".mp3")
+        print(result["text"])
+        result = result["text"]
+        os.remove(filename)
+        os.remove(filename[:-4] + ".mp3")
+        print("Removed video and audio files")
+        print("Done!")
+        return result
+    except:
+        print("Error transcribing audio to text")
+        return
+    
+
+if __name__ == "__main__":
+    main()

requirements.txt

@@ -1,2 +1,7 @@
 gradio
-whisper
+whisper
+pytube
+moviepy
+pywhisper
+flask
+static-ffmpeg