Refactor: Switch from pipeline to direct model usage to resolve task registry errors

app.py

@@ -1,5 +1,5 @@
 import streamlit as st
-from transformers import pipeline
+from transformers import pipeline, AutoTokenizer, AutoModelForSeq2SeqLM
 from PIL import Image
 import pytesseract
 import pdfplumber
@@ -11,9 +11,11 @@ st.write("Summarize text, images, or PDFs with AI")
 
 @st.cache_resource
 def load_model():
-    return pipeline("summarization", model="t5-small")
+    tokenizer = AutoTokenizer.from_pretrained("t5-small")
+    model = AutoModelForSeq2SeqLM.from_pretrained("t5-small")
+    return tokenizer, model
 
-summarizer = load_model()
+tokenizer, model = load_model()
 
 def extract_text_from_image(image):
     """Extract text from image using OCR"""
@@ -33,9 +35,10 @@ def summarize_text(text):
     """Summarize the given text"""
     if not text.strip():
         return None
-    # input_text = "summarize: " + text[:4000]
-    result = summarizer(text[:4000], max_length=150, min_length=40, do_sample=False)
-    return result[0]["summary_text"]
+    input_text = "summarize: " + text[:4000]
+    inputs = tokenizer(input_text, return_tensors="pt", max_length=512, truncation=True)
+    summary_ids = model.generate(inputs.input_ids, max_length=150, min_length=40, length_penalty=2.0, num_beams=4, early_stopping=True)
+    return tokenizer.decode(summary_ids[0], skip_special_tokens=True)
 
 tab1, tab2, tab3 = st.tabs(["📝 Text", "🖼️ Image", "📄 PDF"])