Update app.py

app.py

@@ -2,66 +2,76 @@ import subprocess, sys
 subprocess.run([sys.executable, "-m", "spacy", "download", "en_core_web_sm"],
                stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
 
-import gradio as gr
-import spacy, torch
+import streamlit as st
+import torch
+import spacy
 from transformers import T5ForConditionalGeneration, AutoTokenizer
 
-MODEL_ID = "Hamzasajjad38/t5-small-qg"
-tokenizer = AutoTokenizer.from_pretrained(MODEL_ID, use_fast=True)
-model     = T5ForConditionalGeneration.from_pretrained(MODEL_ID)
-device    = "cuda" if torch.cuda.is_available() else "cpu"
-model     = model.to(device)
-nlp       = spacy.load("en_core_web_sm")
+st.set_page_config(page_title="Question Generator", page_icon="🧠")
 
-def generate_questions(passage, num_questions, question_type):
+@st.cache_resource
+def load_model():
+    MODEL_ID = "Hamzasajjad38/t5-small-qg"
+    tokenizer = AutoTokenizer.from_pretrained(MODEL_ID, use_fast=True)
+    model     = T5ForConditionalGeneration.from_pretrained(MODEL_ID)
+    device    = "cuda" if torch.cuda.is_available() else "cpu"
+    model     = model.to(device)
+    nlp       = spacy.load("en_core_web_sm")
+    return tokenizer, model, device, nlp
+
+tokenizer, model, device, nlp = load_model()
+
+st.title("🧠 Automatic Question Generator")
+st.markdown("**Fine-tuned T5-small on SQuAD 1.1** — paste any passage to generate questions.")
+
+passage     = st.text_area("Input Passage", height=200,
+                            placeholder="Paste any educational paragraph here...")
+num_q       = st.slider("Number of Questions", 1, 5, 3)
+q_type      = st.radio("Question Type", ["Short answer", "True / False"])
+gen_clicked = st.button("⚡ Generate Questions", type="primary")
+
+if gen_clicked:
     if not passage or len(passage.strip()) < 30:
-        return "Please enter a longer passage (at least 30 characters)."
-    doc        = nlp(passage)
-    candidates = [e.text for e in doc.ents]
-    if len(candidates) < num_questions:
-        candidates += [c.text for c in doc.noun_chunks]
-    seen, unique = set(), []
-    for c in candidates:
-        cl = c.lower().strip()
-        if cl not in seen and 1 < len(c.split()) <= 5:
-            seen.add(cl)
-            unique.append(c)
-    candidates = unique[:int(num_questions)]
-    if not candidates:
-        return "Could not extract answer candidates. Try a more detailed passage."
-    lines = []
-    for i, answer in enumerate(candidates):
-        highlighted = passage.replace(answer, f"<hl> {answer} <hl>", 1)
-        inp  = tokenizer(f"generate question: {highlighted}",
-                         return_tensors="pt", max_length=512,
-                         truncation=True).to(device)
-        out  = model.generate(**inp, max_new_tokens=64,
-                              num_beams=4, early_stopping=True)
-        q    = tokenizer.decode(out[0], skip_special_tokens=True)
-        if question_type == "True / False":
-            q = f"True or False: {q.rstrip('?')}?"
-        lines.append(f"Q{i+1}: {q}")
-        lines.append(f"   Answer: {answer}\n")
-    return "\n".join(lines)
+        st.warning("Please enter a longer passage.")
+    else:
+        with st.spinner("Generating questions..."):
+            doc        = nlp(passage)
+            candidates = [e.text for e in doc.ents]
+            if len(candidates) < num_q:
+                candidates += [c.text for c in doc.noun_chunks]
+            seen, unique = set(), []
+            for c in candidates:
+                cl = c.lower().strip()
+                if cl not in seen and 1 < len(c.split()) <= 5:
+                    seen.add(cl)
+                    unique.append(c)
+            candidates = unique[:num_q]
+
+            if not candidates:
+                st.error("Could not extract candidates. Try a more detailed passage.")
+            else:
+                results = []
+                for i, answer in enumerate(candidates):
+                    highlighted = passage.replace(answer, f"<hl> {answer} <hl>", 1)
+                    inp = tokenizer(f"generate question: {highlighted}",
+                                    return_tensors="pt", max_length=512,
+                                    truncation=True).to(device)
+                    out = model.generate(**inp, max_new_tokens=64,
+                                        num_beams=4, early_stopping=True)
+                    q   = tokenizer.decode(out[0], skip_special_tokens=True)
+                    if q_type == "True / False":
+                        q = f"True or False: {q.rstrip('?')}?"
+                    results.append((i+1, q, answer))
 
-demo = gr.Interface(
-    fn=generate_questions,
-    inputs=[
-        gr.Textbox(lines=8, label="Input Passage",
-                   placeholder="Paste any educational paragraph here..."),
-        gr.Slider(1, 5, value=3, step=1, label="Number of Questions"),
-        gr.Radio(["Short answer", "True / False"],
-                 value="Short answer", label="Question Type"),
-    ],
-    outputs=gr.Textbox(lines=14, label="Generated Questions"),
-    title="Automatic Question Generator",
-    description="Fine-tuned T5-small on SQuAD 1.1 — paste any passage to generate questions.",
-    examples=[
-        ["Photosynthesis is a process used by plants to convert light energy into chemical energy stored in glucose. It occurs inside chloroplasts using chlorophyll to absorb sunlight.", 3, "Short answer"],
-        ["The Industrial Revolution began in Britain in the late 18th century. The steam engine invented by James Watt transformed manufacturing and transportation across the country.", 3, "Short answer"],
-        ["Machine learning is a subset of artificial intelligence where systems learn from data to improve performance without being explicitly programmed.", 2, "True / False"],
-    ],
-    allow_flagging="never",
-)
+                st.success(f"Generated {len(results)} questions!")
+                output_text = ""
+                for i, q, ans in results:
+                    st.markdown(f"**Q{i}: {q}**")
+                    st.markdown(f"Answer: `{ans}`")
+                    st.divider()
+                    output_text += f"Q{i}: {q}\nAnswer: {ans}\n\n"
 
-demo.launch()
+                st.download_button("📄 Download as TXT",
+                                   data=output_text,
+                                   file_name="questions.txt",
+                                   mime="text/plain")