Updating to distilbert-base-uncased-finetuned-sst-2-english

app.py

@@ -19,19 +19,16 @@ def anomalies_detector(logs: str) -> list[tuple[int, str]]:
     Returns:
         list[tuple[int, str]]: List of tuples containing (line_number, anomalous_text)
     """
-    # Initialize the text classification pipeline with a model specialized in log analysis
+    # Initialize the text classification pipeline with a model specialized in text classification
     classifier = pipeline(
         "text-classification",
-        model="microsoft/codebert-base",  # Using CodeBERT which is better for technical text
+        model="distilbert-base-uncased-finetuned-sst-2-english",  # Using a model fine-tuned for sentiment/classification
         top_k=2  # Get both normal and anomalous probabilities
     )
-    if logs == "":
-        return []
+    
     # Split logs into lines
     log_lines = logs.split('\n')
     anomalies = []
-    if len(log_lines) == 0:
-        return []
     
     # Process each line
     for line_num, line in enumerate(log_lines, 1):
@@ -42,9 +39,9 @@ def anomalies_detector(logs: str) -> list[tuple[int, str]]:
         results = classifier(line)
         
         # Check if the line is classified as anomalous
-        # CodeBERT returns probabilities for both classes
+        # The model returns probabilities for both classes
         for result in results:
-            if result['label'] == 'LABEL_1' and result['score'] > 0.7:
+            if result['label'] == 'NEGATIVE' and result['score'] > 0.7:  # NEGATIVE indicates potential anomaly
                 anomalies.append((line_num, line))
                 break