Create app.py

app.py

@@ -0,0 +1,43 @@
+import gradio as gr
+from transformers import AutoTokenizer, AutoModelForSequenceClassification
+import torch
+import nltk
+
+nltk.download("punkt")
+
+# Load model
+MODEL = "roberta-base-openai-detector"
+tokenizer = AutoTokenizer.from_pretrained(MODEL)
+model = AutoModelForSequenceClassification.from_pretrained(MODEL)
+
+def detect_ai(text):
+    sentences = nltk.sent_tokenize(text)
+    results = []
+
+    for sent in sentences:
+        inputs = tokenizer(sent, return_tensors="pt", truncation=True, max_length=512)
+        with torch.no_grad():
+            outputs = model(**inputs)
+            probs = torch.softmax(outputs.logits, dim=1)
+        ai_score = float(probs[1])
+        results.append({"sentence": sent, "ai_score": ai_score})
+
+    # Build highlighted HTML
+    highlighted = ""
+    for r in results:
+        color = f"rgba(255,0,0,{r['ai_score']})"  # more red = more AI-like
+        highlighted += f"<span style='background-color:{color}; padding:2px'>{r['sentence']} </span>"
+
+    return highlighted, results
+
+with gr.Blocks() as demo:
+    gr.Markdown("## 🤖 AI Detector (like ZeroGPT)")
+    gr.Markdown("Paste your text below. Redder highlights = more AI-like.")
+    input_text = gr.Textbox(lines=8, placeholder="Enter text here...")
+    output_html = gr.HTML()
+    output_json = gr.JSON()
+    run_btn = gr.Button("Detect AI")
+
+    run_btn.click(detect_ai, inputs=input_text, outputs=[output_html, output_json])
+
+demo.launch()