Update app.py

app.py

@@ -1,25 +1,29 @@
-Python 3.12.10 (tags/v3.12.10:0cc8128, Apr  8 2025, 12:21:36) [MSC v.1943 64 bit (AMD64)] on win32
-Enter "help" below or click "Help" above for more information.
 import gradio as gr
 from transformers import pipeline
 
 # ---------------------------
 # Load Models
 # ---------------------------
+
+# Claim Extraction
 claim_model_name = "microsoft/deberta-v3-base-zeroshot-v1.1"
-claim_classifier = pipeline("zero-shot-classification", model=claim_model_name, device=0)
+claim_classifier = pipeline("zero-shot-classification", model=claim_model_name)
 claim_labels = ["factual claim", "opinion", "personal anecdote", "other"]
 
+# AI-Generated Text Detection
 ai_detect_model_name = "roberta-base-openai-detector"
-ai_detector = pipeline("text-classification", model=ai_detect_model_name, device=0)
+ai_detector = pipeline("text-classification", model=ai_detect_model_name)
 
+# Fact-Checking (NLI)
 nli_model_name = "valhalla/distilbart-mnli-12-3"
-nli_pipeline = pipeline("text-classification", model=nli_model_name, tokenizer=nli_model_name, device=0)
+nli_pipeline = pipeline("text-classification", model=nli_model_name, tokenizer=nli_model_name)
 
 # ---------------------------
 # Functions
 # ---------------------------
+
 def extract_claims(page_text):
+    """Return top 5 factual claims from page text"""
     sentences = [s.strip() for s in page_text.split(".") if len(s.strip()) > 5]
     results = []
     for s in sentences:
@@ -29,6 +33,7 @@ def extract_claims(page_text):
     return results[:5]
 
 def detect_ai(texts):
+    """Detect AI-generated text for a list of strings"""
     if isinstance(texts, str):
         texts = [texts]
     results = []
@@ -38,56 +43,81 @@ def detect_ai(texts):
     return results
 
 def fact_check(claims, evidence_text):
+    """Perform fact-checking using NLI: ENTAILMENT / CONTRADICTION / NEUTRAL"""
     if isinstance(claims, str):
         claims = [claims]
     results = []
     for c in claims:
-        out = nli_pipeline(hypothesis=c, sequence_pair=evidence_text)
+        # Combine claim + evidence for NLI input
+        nli_input = f"{c} </s></s> {evidence_text}"
+        out = nli_pipeline(nli_input)
         results.append({"claim": c, "label": out[0]["label"], "score": round(out[0]["score"], 3)})
     return results
 
 # ---------------------------
 # Unified Predict Function
 # ---------------------------
+
 def predict(page_text="", selected_text="", evidence_text=""):
     """
     1. Extract top 5 claims from page_text
     2. Run AI Detection on claims + selected_text
-    3. Run Fact-Checking on claims + evidence_text if provided
+    3. Run Fact-Checking on claims + selected_text if evidence provided
     """
-    # Extract claims
+    # 1️⃣ Extract claims
     claims = extract_claims(page_text) if page_text else []
-... 
-...     # Combine claims + selected text for AI detection
-...     ai_input = claims.copy()
-...     if selected_text:
-...         ai_input.append(selected_text)
-...     ai_results = detect_ai(ai_input) if ai_input else []
-... 
-...     # Fact-checking: only if evidence is provided
-...     fc_results = fact_check(claims + ([selected_text] if selected_text else []), evidence_text) if evidence_text else []
-... 
-...     return {
-...         "claims": claims,
-...         "ai_detection": ai_results,
-...         "fact_checking": fc_results
-...     }
-... 
-... # ---------------------------
-... # Gradio UI
-... # ---------------------------
-... with gr.Blocks() as demo:
-...     gr.Markdown("## EduShield AI Backend - Predict API & UI")
-... 
-...     page_text_input = gr.Textbox(label="Full Page Text", lines=10, placeholder="Paste page text here...")
-...     selected_text_input = gr.Textbox(label="Selected Text", lines=5, placeholder="Paste selected text here...")
-...     evidence_input = gr.Textbox(label="Evidence Text", lines=5, placeholder="Paste evidence text here...")
-...     predict_btn = gr.Button("Run Predict")
-...     output_json = gr.JSON(label="Predict Results")
-...     predict_btn.click(predict, inputs=[page_text_input, selected_text_input, evidence_input], outputs=output_json)
-... 
-... # ---------------------------
-... # Launch
-... # ---------------------------
-... if __name__ == "__main__":
-...     demo.launch(server_name="0.0.0.0")
+
+    # 2️⃣ AI Detection (claims + selected text)
+    ai_input = claims.copy()
+    if selected_text:
+        ai_input.append(selected_text)
+    ai_results = detect_ai(ai_input) if ai_input else []
+
+    # 3️⃣ Fact-Checking (claims + selected text)
+    fc_input = claims.copy()
+    if selected_text:
+        fc_input.append(selected_text)
+    fc_results = fact_check(fc_input, evidence_text) if evidence_text else []
+
+    return {
+        "claims": claims,
+        "ai_detection": ai_results,
+        "fact_checking": fc_results
+    }
+
+# ---------------------------
+# Gradio UI
+# ---------------------------
+
+with gr.Blocks() as demo:
+    gr.Markdown("## EduShield AI Backend - Predict API & UI")
+    
+    with gr.Tab("Predict"):
+        page_text_input = gr.Textbox(label="Full Page Text", lines=10, placeholder="Paste page text here...")
+        selected_text_input = gr.Textbox(label="Selected Text", lines=5, placeholder="Paste selected text here...")
+        evidence_input = gr.Textbox(label="Evidence Text", lines=5, placeholder="Paste evidence text here...")
+        predict_btn = gr.Button("Run Predict")
+        output_json = gr.JSON(label="Predict Results")
+        predict_btn.click(
+            fn=predict,
+            inputs=[page_text_input, selected_text_input, evidence_input],
+            outputs=output_json
+        )
+    
+    with gr.Tab("Instructions"):
+        gr.Markdown("""
+        ### How to Use EduShield AI Backend
+        1. **Full Page Text**: Paste all visible text from a page. The system will extract the top 5 factual claims.
+        2. **Selected Text**: Paste any specific text to detect AI-generation and/or fact-check.
+        3. **Evidence Text**: Paste evidence against which to fact-check the claims.
+        4. Click **Run Predict** to get results:
+           - **Claims**: Top factual statements from the page
+           - **AI Detection**: Label + probability for AI-generated content
+           - **Fact-Checking**: ENTAILMENT / CONTRADICTION / NEUTRAL with score
+        """)
+
+# ---------------------------
+# Launch
+# ---------------------------
+if __name__ == "__main__":
+    demo.launch()