final interface design

app.py

@@ -1,36 +1,42 @@
-import streamlit as st 
-import streamlit as st
+
+from flask import Flask, render_template, request, jsonify
+from transformers import AutoTokenizer, AutoModelForSequenceClassification
 import torch
 
-from transformers import (
-    DistilBertForSequenceClassification,
-    DistilBertTokenizerFast
-)
+app = Flask(__name__)
+
 MODEL_PATH = "AI-MODEL-FINGERPRINTING/notebooks/saved_distilbert_model"
 
-tokenizer = DistilBertTokenizerFast.from_pretrained(MODEL_PATH)
+print("Loading model...")
 
-model = DistilBertForSequenceClassification.from_pretrained(
-    MODEL_PATH
-)
+tokenizer = AutoTokenizer.from_pretrained(MODEL_PATH)
+model = AutoModelForSequenceClassification.from_pretrained(MODEL_PATH)
 
 model.eval()
 
-st.title("AI Model Fingerprinter")
-
-text = st.text_area(
-    "Paste text here",
-    height=250
-)
-model_names = [
-    "Claude",
-    "Gemini",
-    "Grok",
-    "Mistral",
-    "OpenAI"
+MODEL_NAMES = [
+    "claude",
+    "gemini",
+    "groq",
+    "mistral",
+    "openai"
 ]
 
-if st.button("Analyze"):
+
+@app.route("/")
+def home():
+    return render_template("index.html")
+
+
+@app.route("/predict", methods=["POST"])
+def predict():
+
+    data = request.get_json()
+
+    text = data.get("text", "").strip()
+
+    if not text:
+        return jsonify({"error": "No text provided"})
 
     inputs = tokenizer(
         text,
@@ -42,22 +48,22 @@ if st.button("Analyze"):
 
     with torch.no_grad():
         outputs = model(**inputs)
-    
+        probs = torch.softmax(outputs.logits, dim=1)[0]
 
-    probs = torch.softmax(outputs.logits, dim=1)
-    st.write("Confidence Scores")
+    pred_idx = torch.argmax(probs).item()
 
-    for i, name in enumerate(model_names):
-        st.write(
-        f"{name}: {probs[0][i].item():.4f}"
-    )
+    prediction = MODEL_NAMES[pred_idx]
 
-    pred_class = torch.argmax(probs, dim=1).item()
+    scores = {
+        MODEL_NAMES[i]: round(float(probs[i]), 4)
+        for i in range(len(MODEL_NAMES))
+    }
 
-   
+    return jsonify({
+        "prediction": prediction,
+        "scores": scores
+    })
 
-    predicted_model = model_names[pred_class]
 
-    st.success(
-    f"Predicted Model: {predicted_model}"
-    )
+if __name__ == "__main__":
+    app.run(debug=True)

data/prompts/yu.py

@@ -1,186 +0,0 @@
-from dotenv import load_dotenv
-import os
-import pandas as pd
-import time
-
-import google.generativeai as genai
-
-# ==========================================
-# LOAD ENV VARIABLES
-# ==========================================
-
-load_dotenv()
-
-# ==========================================
-# CONFIGURE GEMINI
-# ==========================================
-
-genai.configure(
-    api_key=os.getenv("GEMINI_API_KEY").strip()
-)
-
-# ==========================================
-# LOAD PROMPT BANK
-# ==========================================
-
-df = pd.read_csv(
-    "AI-MODEL-FINGERPRINTING/src/data_collection/prompt_bank.csv"
-)
-
-# ==========================================
-# FIX COLUMN NAMES
-# ==========================================
-
-df.columns = df.columns.str.strip()
-
-print(df.columns)
-
-# ==========================================
-# SELECT PROMPT RANGE
-# ==========================================
-
-# Example:
-# 0:20   -> prompts 1 to 20
-# 20:40  -> prompts 21 to 40
-# 40:60  -> prompts 41 to 60
-
-df = df.iloc[0:20]
-
-# ==========================================
-# OUTPUT FILE
-# ==========================================
-
-
-output_file = "AI-MODEL-FINGERPRINTING/src/data_collection/gemini_response.csv"
-
-# ==========================================
-# LOAD OLD RESPONSES IF FILE EXISTS
-# ==========================================
-
-if os.path.exists(output_file):
-
-    old_df = pd.read_csv(output_file)
-
-    completed_ids = set(old_df["prompt_id"])
-
-    print(f"\nAlready completed: {len(completed_ids)} prompts")
-
-else:
-
-    old_df = pd.DataFrame()
-
-    completed_ids = set()
-
-# ==========================================
-# GEMINI RESPONSE FUNCTION
-# ==========================================
-
-def generate_gemini_response(prompt):
-
-    try:
-
-        model = genai.GenerativeModel(
-            "models/gemini-2.5-flash"
-        )
-
-        response = model.generate_content(
-            prompt
-        )
-
-        return response.text
-
-    except Exception as e:
-
-        print(f"\nGemini Error: {e}")
-
-        return None
-
-# ==========================================
-# RESPONSE GENERATION LOOP
-# ==========================================
-
-results = []
-
-for index, row in df.iterrows():
-
-    prompt_id = row["PROMPT_ID"]
-
-    # ======================================
-    # SKIP COMPLETED PROMPTS
-    # ======================================
-
-    if prompt_id in completed_ids:
-
-        print(f"\nSkipping {prompt_id} (already done)")
-
-        continue
-
-    category = row["category"]
-
-    prompt = row["PROMPT"]
-
-    print(f"\nGenerating response for {prompt_id}...")
-
-    generated_text = generate_gemini_response(
-        prompt
-    )
-
-    # ======================================
-    # HANDLE FAILED GENERATION
-    # ======================================
-
-    if generated_text is None:
-
-        print(f"\nFailed for {prompt_id}")
-
-        continue
-
-    # ======================================
-    # STORE RESULT
-    # ======================================
-
-    result = {
-
-        "prompt_id": prompt_id,
-
-        "category": category,
-
-        "model": "gemini",
-
-        "prompt": prompt,
-
-        "response": generated_text
-
-    }
-
-    results.append(result)
-
-    # ======================================
-    # SAVE AFTER EVERY RESPONSE
-    # ======================================
-
-    temp_df = pd.DataFrame(results)
-
-    final_df = pd.concat(
-        [old_df, temp_df],
-        ignore_index=True
-    )
-
-    final_df.to_csv(
-        output_file,
-        index=False
-    )
-
-    print(f"\nSaved {prompt_id}")
-
-    # ======================================
-    # DELAY TO AVOID RATE LIMITS
-    # ======================================
-
-    time.sleep(3)
-
-# ==========================================
-# FINAL MESSAGE
-# ==========================================
-
-print("\nGemini response generation completed.")

requirement.txt

@@ -11,4 +11,6 @@ datasets
 accelerate
 evaluate
 sentencepiece
-shap
+shap
+flask
+safetensors

static/script.js

@@ -0,0 +1,80 @@
+document.addEventListener("DOMContentLoaded", () => {
+
+    const form = document.querySelector("form");
+
+    form.addEventListener("submit", async (e) => {
+
+        e.preventDefault();
+
+        const text = document.querySelector("textarea").value;
+
+        let resultDiv = document.getElementById("result");
+
+        if (!resultDiv) {
+            resultDiv = document.createElement("div");
+            resultDiv.id = "result";
+            document.querySelector(".container").appendChild(resultDiv);
+        }
+
+        resultDiv.innerHTML = `
+            <div class="result-box">
+                <p> Analyzing...</p>
+            </div>
+        `;
+
+        try {
+
+            const response = await fetch("/predict", {
+                method: "POST",
+                headers: {
+                    "Content-Type": "application/json"
+                },
+                body: JSON.stringify({
+                    text: text
+                })
+            });
+
+            const data = await response.json();
+
+            let html = `
+                <div class="result-box">
+
+                    <h2>Prediction</h2>
+
+                    <p class="prediction">
+                         ${data.prediction.toUpperCase()}
+                    </p>
+
+                    <h3>Confidence Scores</h3>
+            `;
+
+            for (const [model, score] of Object.entries(data.scores)) {
+
+                html += `
+                    <p>
+                        <strong>${model.toUpperCase()}</strong> :
+                        ${score}
+                    </p>
+                `;
+            }
+
+            html += `
+                </div>
+            `;
+
+            resultDiv.innerHTML = html;
+
+        } catch (error) {
+
+            resultDiv.innerHTML = `
+                <div class="result-box">
+                    <p>Error while analyzing text.</p>
+                </div>
+            `;
+
+            console.error(error);
+        }
+
+    });
+
+});

static/style.css

@@ -0,0 +1,66 @@
+body {
+    background-color: #0f172a;
+    color: #f8fafc;
+    font-family: Arial, sans-serif;
+    margin: 0;
+    padding: 0;
+}
+
+.container {
+    width: 80%;
+    max-width: 900px;
+    margin: 50px auto;
+    text-align: center;
+}
+
+h1 {
+    font-size: 3rem;
+    margin-bottom: 10px;
+    color:#60a5fa;
+}
+
+.subtitle {
+    color: #94a3b8;
+    margin-bottom: 30px;
+}
+
+textarea {
+    width: 100%;
+    height: 220px;
+    padding: 15px;
+    border-radius: 12px;
+    border: none;
+    background: #1e293b;
+    color: white;
+    font-size: 16px;
+    resize: none;
+}
+
+button {
+    margin-top: 20px;
+    padding: 12px 30px;
+    border: none;
+    border-radius: 10px;
+    background: #3b82f6;
+    color: white;
+    font-size: 18px;
+    cursor: pointer;
+}
+
+button:hover {
+    background: #2563eb;
+}
+
+.result-box {
+    margin-top: 30px;
+    padding: 20px;
+    background: #1e293b;
+    border-radius: 12px;
+}
+
+.prediction {
+    font-size: 32px;
+    color: #22c55e;
+    font-weight: bold;
+    letter-spacing: 2px;
+}

templates/index.html

@@ -0,0 +1,41 @@
+<!DOCTYPE html>
+<html>
+<head>
+    <title>AI origin detector</title>
+    <link rel="stylesheet" href="{{ url_for('static', filename='style.css') }}">
+</head>
+
+<body>
+
+<div class="container">
+
+    <h1> AI Origin Detector</h1>
+    <p class="subtitle">
+        Identify which AI model generated a piece of text.
+    </p>
+
+    <form method="POST">
+        <textarea
+            name="text"
+            placeholder="Paste AI generated text here..."
+        ></textarea>
+
+        <br>
+
+        <button type="submit">
+            Analyze
+        </button>
+    </form>
+
+    {% if prediction %}
+    <div class="result-box">
+        <h2>Prediction</h2>
+        <p class="prediction">{{ prediction }}</p>
+    </div>
+    {% endif %}
+
+</div>
+<script src="{{ url_for('static', filename='script.js') }}"></script>
+
+</body>
+</html>