Rename ai_text_detector.py to model_classifier.py

ai_text_detector.py → model_classifier.py

@@ -6,7 +6,6 @@ from tokenizers.normalizers import Sequence, Replace, Strip, NFKC
 from tokenizers import Regex
 import os
 
-
 device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
 
 # Updated model paths with the correct model 1 location
@@ -74,7 +73,7 @@ def clean_text(text: str) -> str:
     text = re.sub(r'\s+([,.;:?!])', r'\1', text)
     return text
 
-newline_to_space  = Replace(Regex(r'\s*\n\s*'), " ")
+newline_to_space = Replace(Regex(r'\s*\n\s*'), " ")
 join_hyphen_break = Replace(Regex(r'(\w+)[--]\s*\n\s*(\w+)'), r"\1\2")
 
 tokenizer.backend_tokenizer.normalizer = Sequence([
@@ -85,6 +84,11 @@ tokenizer.backend_tokenizer.normalizer = Sequence([
 ])
 
 def classify_text(text):
+    """
+    Classify text using ModernBERT ensemble
+    Author: deveshpunjabi
+    Date: 2025-01-15 07:07:03 UTC
+    """
     cleaned_text = clean_text(text)
     if not text.strip():
         return "⚠️ Please enter some text to analyze"
@@ -115,9 +119,7 @@ def classify_text(text):
         result_message = f"""
 ### 🟢 **Human Written**
 **Confidence: {human_prob:.2f}%**
-
 This text appears to be written by a human.
-
 ---
 **Analysis Details:**
 - Human probability: {human_prob:.2f}%
@@ -128,11 +130,8 @@ This text appears to be written by a human.
         result_message = f"""
 ### 🔴 **AI Generated**
 **Confidence: {ai_total_prob:.2f}%**
-
 **Most likely source: {ai_argmax_model}**
-
 This text appears to be generated by an AI model.
-
 ---
 **Analysis Details:**
 - Human probability: {human_prob:.2f}%