Update app.py

app.py

@@ -147,13 +147,13 @@ def compute_features(text):
     punctuation_ratio = punctuation_count / (len(text) + 1e-6)
     avg_word_length = sum(len(w) for w in words) / (word_count if word_count else 1)
 
-    return np.array([[word_count, unique_word_ratio, repeat_rate, punctuation_ratio, avg_word_length]])
+    return [[word_count, unique_word_ratio, repeat_rate, punctuation_ratio, avg_word_length]]
 
 # ---------------- 使用 scaler ----------------
 def transform_features(feat):
     if scaler is None:
         return feat  # 如果 scaler 沒載入，就直接回傳原始特徵
-    return scaler.transform(feat)
+    return scaler.transform(feat).tolist()  # 轉成 list，避免使用 NumPy
 
 # ---------------- 生成解釋 ----------------
 def explain_prediction(text):
@@ -169,7 +169,7 @@ def explain_prediction(text):
         seq = vectorized_layer([text])
         seq = tf.keras.utils.pad_sequences(seq, maxlen=50, padding='pre')
 
-        # 轉成 Tensor
+        # 轉成 TensorFlow tensor
         seq = tf.convert_to_tensor(seq)
         feat = tf.convert_to_tensor(feat, dtype=tf.float32)
 
@@ -202,4 +202,4 @@ iface = gr.Interface(
     description="輸入文章，模型會判斷是 AI 或人類撰寫，並給出機率與判斷依據"
 )
 
-iface.launch()
+iface.launch()