Update app.py

app.py

@@ -1,85 +1,55 @@
-import re
 import gradio as gr
 from youtube_transcript_api import YouTubeTranscriptApi
 from transformers import pipeline
+import re
 
-# Lazy load models (critical for HF Spaces)
-summarizer = None
-qa_pipeline = None
-
-def load_models():
-    global summarizer, qa_pipeline
-    if summarizer is None:
-        summarizer = pipeline(
-            task="text2text-generation",
-            model="sshleifer/distilbart-cnn-12-6"
-        )
-    if qa_pipeline is None:
-        qa_pipeline = pipeline(
-            task="question-answering",
-            model="distilbert-base-cased-distilled-squad"
-        )
+# Load summarizer (lightweight for Spaces)
+summarizer = pipeline("summarization", model="facebook/bart-large-cnn")
 
 def extract_video_id(url):
-    regex = r"(?:v=|youtu\.be/)([a-zA-Z0-9_-]{11})"
+    regex = r"(?:v=|\/)([0-9A-Za-z_-]{11}).*"
     match = re.search(regex, url)
     return match.group(1) if match else None
 
-def get_transcript(url):
-    video_id = extract_video_id(url)
-    if not video_id:
-        return None, "❌ Invalid YouTube URL"
+def get_transcript(video_id):
     try:
         transcript = YouTubeTranscriptApi.get_transcript(video_id)
-        text = " ".join([t["text"] for t in transcript])
-        return text, None
-    except Exception as e:
-        return None, f"❌ Transcript Error: {str(e)}"
-
-def chunk_text(text, size=400):
-    words = text.split()
-    for i in range(0, len(words), size):
-        yield " ".join(words[i:i+size])
+        text = " ".join([t['text'] for t in transcript])
+        return text
+    except:
+        return None
 
 def summarize_text(text):
-    load_models()
-    result = []
-    for chunk in chunk_text(text):
-        out = summarizer(chunk, max_length=100, min_length=30, do_sample=False)
-        result.append(out[0]["generated_text"])
-    return " ".join(result)
-
-def answer_question(context, question):
-    load_models()
-    best = ("", 0)
-    for chunk in chunk_text(context):
-        try:
-            res = qa_pipeline(question=question, context=chunk)
-            if res["score"] > best[1]:
-                best = (res["answer"], res["score"])
-        except:
-            continue
-    return best[0] if best[0] else "❌ No answer found"
-
-def process(url, question):
-    text, err = get_transcript(url)
-    if err:
-        return err, ""
-    summary = summarize_text(text)
-    answer = answer_question(text, question) if question.strip() else ""
-    return summary, answer
-
-demo = gr.Interface(
-    fn=process,
-    inputs=[
-        gr.Textbox(label="YouTube URL"),
-        gr.Textbox(label="Ask a question (optional)")
-    ],
-    outputs=[
-        gr.Textbox(label="Summary"),
-        gr.Textbox(label="Answer")
-    ],
-    title="YouTube Video Summarizer + Chat"
+    max_chunk = 1000
+    chunks = [text[i:i+max_chunk] for i in range(0, len(text), max_chunk)]
+    
+    summary = ""
+    for chunk in chunks:
+        result = summarizer(chunk, max_length=130, min_length=30, do_sample=False)
+        summary += result[0]['summary_text'] + " "
+    
+    return summary
+
+def process_video(url):
+    video_id = extract_video_id(url)
+    if not video_id:
+        return "Invalid YouTube URL"
+
+    transcript = get_transcript(video_id)
+    if not transcript:
+        return "Could not fetch transcript"
+
+    summary = summarize_text(transcript)
+    return summary
+
+# Gradio UI
+iface = gr.Interface(
+    fn=process_video,
+    inputs=gr.Textbox(label="Enter YouTube URL"),
+    outputs=gr.Textbox(label="Summary"),
+    title="YouTube Video Summarizer",
+    description="Paste a YouTube link to get a summarized transcript."
 )
 
-demo.launch()
+if __name__ == "__main__":
+    iface.launch()