Update app.py

app.py

@@ -1,12 +1,8 @@
 import streamlit as st
 from transformers import pipeline
 
-# Cache the summarizer to avoid reloading on every interaction
-@st.cache_resource
-def load_summarizer():
-    return pipeline("summarization", model="facebook/bart-large-cnn")
-
-summarizer = load_summarizer()
+# Initialize the summarization pipeline
+summarizer = pipeline("summarization", model="facebook/bart-large-cnn")
 
 # Streamlit UI setup
 st.title("📝 Text Summarization App")
@@ -53,21 +49,39 @@ else:  # Long
 # Display the selected range for feedback
 st.write(f"Selected Summary Length: {'Short' if summary_length == 1 else 'Medium' if summary_length == 2 else 'Long'}")
 
+# Function to split text into manageable chunks
+def chunk_text(text, max_chunk_size=1024):
+    tokens = text.split()
+    chunks = []
+    current_chunk = []
+
+    for token in tokens:
+        current_chunk.append(token)
+        if len(' '.join(current_chunk)) > max_chunk_size:
+            chunks.append(' '.join(current_chunk[:-1]))
+            current_chunk = [token]
+    chunks.append(' '.join(current_chunk))  # Add the final chunk
+    return chunks
+
 # Button to trigger summarization
 if st.button("Summarize"):
     if user_input.strip():  # Ensure there's input before summarizing
         try:
-            # Generate summary using the model with custom min and max length
-            summarized_text = summarizer(user_input, 
-                                         max_length=max_len, 
-                                         min_length=min_len, 
-                                         length_penalty=2.0, 
-                                         num_beams=4, 
-                                         early_stopping=True)[0]['summary_text']
+            # Split the input into chunks
+            text_chunks = chunk_text(user_input)
+
+            # Summarize each chunk separately
+            summaries = []
+            for chunk in text_chunks:
+                summary = summarizer(chunk, max_length=max_len, min_length=min_len, length_penalty=2.0, num_beams=4, early_stopping=True)[0]['summary_text']
+                summaries.append(summary)
+
+            # Combine summaries from all chunks
+            full_summary = " ".join(summaries)
 
             # Display the generated summary
             st.subheader("Summarized Text:")
-            st.write(summarized_text)
+            st.write(full_summary)
 
         except Exception as e:
             st.error(f"An error occurred while summarizing: {e}")