Update src/streamlit_app.py

src/streamlit_app.py

@@ -7,32 +7,32 @@ import random
 from collections import defaultdict
 import json
 
-# Label encoder yükləmə funksiyası
-def load_label_encoder():
-    file_path = os.path.join(os.getcwd(), "best_model", "label_encoder.pkl")
+# Name encoder yükləmə funksiyası
+def load_name_encoder():
+    file_path = os.path.join(os.getcwd(), "best_model", "name_encoder.pkl")
     if not os.path.exists(file_path):
-        st.error(f"Label encoder faylı tapılmadı: {file_path}")
+        st.error(f"Name encoder faylı tapılmadı: {file_path}")
         st.stop()
     with open(file_path, "rb") as f:
-        label_encoder = pickle.load(f)
-    return label_encoder
+        name_encoder = pickle.load(f)
+    return name_encoder
 
 # Model və tokenizer yükləmə
 @st.cache_resource
 def load_model():
-    label_encoder = load_label_encoder()
+    name_encoder = load_name_encoder()
     model_path = os.path.join(os.getcwd(), "best_model")
 
     tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')
     model = BertForSequenceClassification.from_pretrained(
         model_path,
-        num_labels=len(label_encoder.classes_) if hasattr(label_encoder, "classes_") else len(label_encoder)
+        num_labels=len(name_encoder.classes_)
     )
     model.eval()
-    return tokenizer, model, label_encoder
+    return tokenizer, model, name_encoder
 
 # Prediction funksiyası
-def predict_disease(symptoms_text, tokenizer, model, label_encoder):
+def predict_disease(symptoms_text, tokenizer, model, name_encoder):
     symptoms = [s.strip() for s in symptoms_text.split(",") if s.strip()]
     agg_probs = defaultdict(float)
     n_shuffles = 10
@@ -60,13 +60,7 @@ def predict_disease(symptoms_text, tokenizer, model, label_encoder):
     top_3 = sorted(agg_probs.items(), key=lambda x: x[1], reverse=True)[:3]
     results = []
     for idx, prob in top_3:
-        # LabelEncoder obyekti olub-olmamasını yoxla
-        if hasattr(label_encoder, "classes_"):
-            label = label_encoder.classes_[idx]
-        elif isinstance(label_encoder, dict):
-            label = label_encoder.get(idx, f"Unknown label {idx}")
-        else:
-            label = f"Unknown label {idx}"
+        label = name_encoder.classes_[idx]
         results.append({"disease": label, "probability": prob})
 
     return results
@@ -79,13 +73,13 @@ query_params = st.query_params
 is_api_mode = str(query_params.get("api", ["false"])[0]).lower() == "true"
 
 # Model yüklə
-tokenizer, model, label_encoder = load_model()
+tokenizer, model, name_encoder = load_model()
 
 # API mode
 if is_api_mode:
     symptoms = query_params.get("symptoms", [""])[0]
     if symptoms:
-        results = predict_disease(symptoms, tokenizer, model, label_encoder)
+        results = predict_disease(symptoms, tokenizer, model, name_encoder)
         api_response = {
             "status": "success",
             "input": symptoms,
@@ -97,7 +91,6 @@ if is_api_mode:
             "message": "symptoms parameter required"
         }
 
-    # JSON olaraq qaytar (raw)
     st.markdown(
         f"```json\n{json.dumps(api_response, ensure_ascii=False, indent=2)}\n```"
     )
@@ -108,10 +101,7 @@ st.title("🏥 Disease Prediction")
 st.success("Model yükləndi!")
 
 # Debug: Siniflər
-if hasattr(label_encoder, "classes_"):
-    st.write("Available classes:", list(label_encoder.classes_))
-elif isinstance(label_encoder, dict):
-    st.write("Available classes:", list(label_encoder.values()))
+st.write("Available classes:", list(name_encoder.classes_))
 
 # API usage info
 st.markdown("### API İstifadəsi")
@@ -119,6 +109,7 @@ space_url = "https://your-username-your-space-name.hf.space"
 api_example = f"{space_url}/?api=true&symptoms=fever,cough,headache"
 st.code(api_example)
 
+# Form
 with st.form(key="predict_form"):
     text = st.text_area("Simptomları daxil edin (vergüllə ayırın):")
     submit_button = st.form_submit_button(label="Predict")
@@ -127,8 +118,7 @@ if submit_button:
     if not text.strip():
         st.warning("Simptomları daxil edin!")
     else:
-        results = predict_disease(text, tokenizer, model, label_encoder)
+        results = predict_disease(text, tokenizer, model, name_encoder)
         st.subheader("🔍 Nəticələr:")
         for result in results:
             st.write(f"**{result['disease']}** — {result['probability']*100:.2f}%")
-