Update app.py

app.py

@@ -65,6 +65,15 @@ def create_dataset(data, tokenizer, max_length):
     texts = [item.get('prompt', '') for item in data]
     labels = [item.get('label', -1) for item in data]
     
+    # Debugging: Print out labels to check for invalid values
+    print(f"Labels before adjustment: {labels}")
+    
+    # Ensure all labels are within the valid range
+    labels = [label if 0 <= label < num_labels else 0 for label in labels]
+    
+    # Debugging: Print out adjusted labels
+    print(f"Labels after adjustment: {labels}")
+    
     encodings = tokenizer(texts, truncation=True, padding='max_length', max_length=max_length)
     dataset = Dataset.from_dict({
         'input_ids': encodings['input_ids'],
@@ -73,6 +82,7 @@ def create_dataset(data, tokenizer, max_length):
     })
     return dataset
 
+
 def split_data(data, test_size=0.2):
     if not data:
         raise ValueError("Data is empty, cannot split.")