Update app.py

app.py

@@ -1,31 +1,30 @@
-
 import streamlit as st
-from youtube_transcript_api import YouTubeTranscriptApi
+from pytube import YouTube
 from crewai import Agent, Task, Crew
 from transformers import pipeline
 import torch
-
-# Hugging Face summarizer setup
-summarizer = pipeline("summarization", model="facebook/bart-large-cnn")
+import whisper
 
 st.title("AI Blog Writer from YouTube")
 
-# YouTube URL Input
-video_url = st.text_input("Enter YouTube Video URL:")
+# Whisper model setup (Base Model for better speed)
+whisper_model = whisper.load_model("base")
 
-def get_video_id(url):
-    """Extract YouTube video ID from the URL."""
-    import re
-    match = re.search(r"(?<=v=)[^&]+", url)
-    if match:
-        return match.group(0)
-    else:
-        st.error("Invalid YouTube URL. Please try again.")
-        return None
+def download_audio(video_url):
+    """Downloads audio from the YouTube video."""
+    yt = YouTube(video_url)
+    audio_stream = yt.streams.filter(only_audio=True).first()
+    audio_path = audio_stream.download(filename="audio.mp4")
+    return audio_path
+
+def transcribe_audio(audio_path):
+    """Transcribes the audio using Whisper AI."""
+    transcription = whisper_model.transcribe(audio_path)
+    return transcription['text']
 
-# CrewAI Agent Setup
+# CrewAI Setup
 def create_crew_agents(transcript_text):
-    research_agent = Agent(name="Research Agent", goal="Extract transcript from videos.")
+    research_agent = Agent(name="Research Agent", goal="Extract video transcript.")
     blog_writer_agent = Agent(name="Blog Writer", goal="Generate blog content.")
     editor_agent = Agent(name="Editor", goal="Refine blog for clarity.")
 
@@ -36,17 +35,21 @@ def create_crew_agents(transcript_text):
     crew = Crew(agents=[research_agent, blog_writer_agent, editor_agent], tasks=[research_task, blog_task, editor_task])
     return crew
 
-# Button Trigger
+# User Input for Video URL
+video_url = st.text_input("Enter YouTube Video URL:")
+
 if st.button("Generate Blog"):
-    video_id = get_video_id(video_url)
-    if video_id:
+    if video_url:
         try:
-            # Fetch Transcript using YouTube API
-            transcript = YouTubeTranscriptApi.get_transcript(video_id)
-            transcript_text = " ".join([item['text'] for item in transcript])
-            st.text_area("Transcript Extracted", transcript_text)
-
-            # CrewAI Agents Collaboration
+            st.info("Downloading audio...")
+            audio_path = download_audio(video_url)
+            st.success("Audio downloaded successfully!")
+            
+            st.info("Transcribing audio using Whisper...")
+            transcript_text = transcribe_audio(audio_path)
+            st.text_area("Transcript Extracted", transcript_text, height=200)
+
+            # Generate Blog with CrewAI
             crew = create_crew_agents(transcript_text)
             results = crew.kickoff()
             st.subheader("Generated Blog")
@@ -54,3 +57,5 @@ if st.button("Generate Blog"):
 
         except Exception as e:
             st.error(f"Error: {e}")
+    else:
+        st.warning("Please enter a valid YouTube URL.")