Update app.py

app.py

@@ -1,26 +1,29 @@
 """
-Governance-GPT Quiz – bucket version (LLM-powered summary)
-Author: Rishabh Sharma – 2025-09-14
-Single-file Gradio Space: 15-question Likert survey → bucket scores → LLM summary + PDF
+Governance-GPT Quiz  ─  LLM-powered bucket report
+Author: Rishabh Sharma · 2025-09-14
+One-file Gradio Space → 15-Q Likert survey → bucket scores → FLAN-T5 summary → PDF
 """
 
-import os, tempfile, datetime, warnings
+import datetime, tempfile, warnings
 import gradio as gr
 import pandas as pd
-from fpdf import FPDF               # pure-python PDF
-from transformers import pipeline    # 🔹 NEW: LLM summariser
+from fpdf import FPDF                 # pure-Python PDF (core Helvetica font)
+from transformers import pipeline     # Hugging Face inference
 
-# ───────────── Initialize LLM once ───────────── #
-# Small, instruction-tuned model that runs comfortably on free CPU Spaces
-warnings.filterwarnings("ignore", category=UserWarning)  # tokenizer warnings
+# ───────────────────────────────────────────────────────────
+# 1.  Load a lightweight model once (≈ 80 M params, CPU-friendly)
+# ───────────────────────────────────────────────────────────
+warnings.filterwarnings("ignore", category=UserWarning)
 summariser = pipeline(
     task="text2text-generation",
     model="google/flan-t5-small",
     tokenizer="google/flan-t5-small",
-    max_new_tokens=220,
+    max_new_tokens=150,
 )
 
-# ───────────── Question bank ───────────── #
+# ───────────────────────────────────────────────────────────
+# 2.  Question bank & buckets
+# ───────────────────────────────────────────────────────────
 QUESTIONS = [
     "Governance framework is documented and communicated across the organisation.",
     "Roles & responsibilities for AI oversight are clearly assigned.",
@@ -59,7 +62,9 @@ TIERS = {
     "Optimized":  (4.51, 5.00),
 }
 
-# ───────────── Helpers ───────────── #
+# ───────────────────────────────────────────────────────────
+# 3.  Helper functions
+# ───────────────────────────────────────────────────────────
 def score_to_tier(avg: float) -> str:
     for tier, (low, high) in TIERS.items():
         if low <= avg <= high:
@@ -67,40 +72,48 @@ def score_to_tier(avg: float) -> str:
     return "Unclassified"
 
 def latin1(txt: str) -> str:
-    return txt.replace("…", "...").encode("latin-1", "replace").decode("latin-1")
+    """Force text into Latin-1 for the core fonts used by FPDF."""
+    return txt.encode("latin-1", "replace").decode("latin-1")
 
 def llm_remediation(product: str, bucket_avgs: dict, overall_tier: str) -> str:
-    """Call Flan-T5 to get a markdown summary with recommendations."""
-    # Craft a concise prompt to stay within max_new_tokens
-    bucket_lines = "\n".join(
-        f"- {b}: {v:.2f}" for b, v in bucket_avgs.items()
-    )
+    """
+    Ask FLAN-T5-small to write a short markdown summary.
+    Uses only ASCII bullets ('-') so the PDF core font can render them.
+    """
+    bucket_lines = "\n".join(f"- {b}: {v:.2f}" for b, v in bucket_avgs.items())
     prompt = (
         "You are an AI governance consultant.\n"
         f"Product: {product}\n"
         f"Overall tier: {overall_tier}\n"
         "Bucket scores (1-5):\n"
         f"{bucket_lines}\n\n"
-        "Write a short markdown summary (≤120 words) with:\n"
-        "• A one-sentence overall assessment\n"
-        "• 3-5 bullet remediation actions, prioritised, referencing bucket names.\n"
-        "Do not add any other sections."
+        "Write a concise markdown summary (≤120 words) with:\n"
+        "- One overall assessment sentence.\n"
+        "- 3-5 bullet remediation actions referencing bucket names.\n"
+        "Return only the summary."
     )
-    try:
-        gen = summariser(prompt)[0]["generated_text"]
-        return gen.strip()
-    except Exception as e:
-        # Fallback in rare failure cases
-        return f"LLM summary unavailable – {e}"
-
-def build_pdf(product: str, bucket_df: pd.DataFrame, overall_avg: float,
-              overall_tier: str, file_path: str, summary_text: str):
+
+    # Generate
+    raw = summariser(prompt)[0]["generated_text"].strip()
+
+    # Some T5 models echo a slice of the prompt. Remove if present.
+    if raw.startswith(prompt[:50]):
+        raw = raw[len(prompt):].strip()
+
+    # Replace any unicode bullets with ASCII "- "
+    summary = raw.replace("•", "- ")
+    return summary if summary else "Summary unavailable."
+
+def build_pdf(product: str, bucket_df: pd.DataFrame,
+              overall_avg: float, overall_tier: str,
+              file_path: str, summary_text: str):
     pdf = FPDF()
     pdf.set_auto_page_break(auto=True, margin=15)
     pdf.add_page()
 
     pdf.set_font("Helvetica", "B", 16)
     pdf.cell(0, 10, latin1(f"AI Governance Maturity Report – {product}"), ln=1, align="C")
+
     pdf.set_font("Helvetica", "", 12)
     pdf.cell(0, 8, f"Generated on {datetime.date.today().isoformat()}", ln=1, align="C")
     pdf.ln(4)
@@ -128,11 +141,14 @@ def build_pdf(product: str, bucket_df: pd.DataFrame, overall_avg: float,
 
     pdf.output(file_path)
 
-# ───────────── Gradio callback ───────────── #
+# ───────────────────────────────────────────────────────────
+# 4.  Gradio callback
+# ───────────────────────────────────────────────────────────
 def generate_report(product_name, *scores):
     product = product_name.strip() or "your product"
     scores = list(scores)
 
+    # Bucket averages
     bucket_avgs = {
         bucket: sum(scores[i] for i in idxs) / len(idxs)
         for bucket, idxs in BUCKETS.items()
@@ -153,13 +169,15 @@ def generate_report(product_name, *scores):
 
     return summary_md, tmp_pdf.name
 
-# ───────────── UI ───────────── #
+# ───────────────────────────────────────────────────────────
+# 5.  Gradio UI
+# ───────────────────────────────────────────────────────────
 with gr.Blocks(title="Governance-GPT Quiz") as demo:
     gr.Markdown(
         """
 # Governance-GPT Quiz  
 Enter your **product / system name**, rate each statement from **1 (Strongly Disagree)** to **5 (Strongly Agree)**,  
-and receive an LLM-generated remediation plan plus PDF bucket report.
+and receive an LLM-generated remediation plan plus a PDF bucket report.
         """
     )