Fix sklearn dependency issue and improve error handling

- Updated scikit-learn version in requirements.txt to >=1.0.0
- Added explicit sklearn import check with helpful error message
- Improved error handling with fallback mode functionality
- Updated interface description to show status when model fails to load
- Standardized naming to 'Medical Case Classifier' across all files

app.py

@@ -3,8 +3,18 @@ import torch
 from transformers import AutoTokenizer, AutoModelForSequenceClassification
 import pickle
 
+# Initialize variables
+model, tokenizer, label_encoder, class_names = None, None, None, []
+
 # Load the saved model, tokenizer, and label encoder
 try:
+    # Check if sklearn is available
+    try:
+        import sklearn
+        print("scikit-learn is available")
+    except ImportError:
+        raise ImportError("scikit-learn is not installed. Please install it using: pip install scikit-learn")
+    
     # Use the correct path where you saved your model
     model_path = "./medical_classifier_model"
     
@@ -29,7 +39,8 @@ try:
 
 except Exception as e:
     print(f"Error loading model components: {e}")
-    # Fallback or exit if loading fails
+    print("The application will run in fallback mode with limited functionality.")
+    # Fallback values when loading fails
     model, tokenizer, label_encoder, class_names = None, None, None, []
 
 def predict_medical_specialty(text):
@@ -85,6 +96,12 @@ examples = [
 num_classes = len(class_names) if class_names else 3
 include_examples = examples if all([model, tokenizer, label_encoder]) else None
 
+# Create appropriate description based on model availability
+if all([model, tokenizer, label_encoder]):
+    description_text = "This application uses a fine-tuned Bio_ClinicalBERT model to predict the medical specialty of a given text."
+else:
+    description_text = "⚠️ Model loading failed (likely missing scikit-learn dependency). Please ensure all requirements are installed. The app is running in fallback mode."
+
 iface = gr.Interface(
     fn=predict_medical_specialty,
     inputs=gr.Textbox(
@@ -94,7 +111,7 @@ iface = gr.Interface(
     ),
     outputs=gr.Label(num_top_classes=num_classes),
     title="Medical Case Classifier",
-    description="This application uses a fine-tuned Bio_ClinicalBERT model to predict the medical specialty of a given text.",
+    description=description_text,
     examples=include_examples
 )
 

requirements.txt

@@ -2,4 +2,4 @@ transformers
 datasets
 torch
 gradio
-scikit-learn
+scikit-learn>=1.0.0