diff --git a/test.py b/test.py
index 1a8475d..9ff7002 100644
--- a/test.py
+++ b/test.py
@@ -1,17 +1,18 @@
 import keras
 from sklearn.metrics import classification_report
 from utils import wav2mfcc, model, get_data
-from keras.utils import to_categorical
+import numpy as np
+import tensorflow as tf
 
 def check_preds(X, y):
 
     trained_model = keras.models.load_model('trained_model.h5')
-    predictions = trained_model.predict_classes(X)
+    predictions = trained_model.predict(X)
+    classes_x = np.argmax(predictions, axis=1)
+    print(classification_report(y, tf.keras.utils.to_categorical(classes_x)))
 
-    print(classification_report(y, to_categorical(predictions)))
 
+if __name__ == '__main__':
+    _, X_test, _, y_test, _ = get_data.get_all()
 
-# if __name__ == '__main__':
-#     _, X_test, _, y_test, _ = get_data.get_all()
-#
-#     check_preds(X_test, y_test)
\ No newline at end of file
+    check_preds(X_test, y_test)
\ No newline at end of file
diff --git a/utils/model.py b/utils/model.py
index adf298f..27e8b9e 100644
--- a/utils/model.py
+++ b/utils/model.py
@@ -28,7 +28,7 @@ def get_cnn_model(input_shape, num_classes):
     model.add(BatchNormalization())
     model.add(Dropout(0.4))
     model.add(Dense(num_classes, activation='softmax'))
-    model.compile(loss=keras.losses.categorical_crossentropy, optimizer=keras.optimizers.Adadelta(), metrics=['accuracy'])
+    model.compile(loss="categorical_crossentropy", optimizer='Adadelta', metrics=['accuracy'])
 
     return model
 
diff --git a/utils/wav2mfcc.py b/utils/wav2mfcc.py
index 01ff482..9e6b5df 100644
--- a/utils/wav2mfcc.py
+++ b/utils/wav2mfcc.py
@@ -1,7 +1,7 @@
 import numpy as np
 import librosa
 import os
-from keras.utils import to_categorical
+import tensorflow as tf
 
 def wav2mfcc(file_path, max_pad_len=20):
     wave, sr = librosa.load(file_path, mono=True, sr=None)
@@ -25,7 +25,7 @@ def get_data():
             label = f.split('_')[0]
             labels.append(label)
 
-    return np.asarray(mfccs), to_categorical(labels)
+    return np.asarray(mfccs), tf.keras.utils.to_categorical(labels)
 
 # if __name__ == '__main__':
 #     mfccs, labels = get_data()