Keras - Validation Loss and Accuracy stuck at 0

Question

I am trying to train a simple 2 layer Fully Connected neural net for Binary Classification in Tensorflow keras. I have split my data into Training and Validation sets with a 80-

Answer 1

• If you use keras instead of tf.keras everything works fine.

• With tf.keras, I even tried validation_data = [X_train, y_train], this also gives zero accuracy.

Here is a demonstration:

model.fit(X_train, y_train, validation_data=[X_train.to_numpy(), y_train.to_numpy()], 
epochs=10, batch_size=64)

Epoch 1/10
8/8 [==============================] - 0s 6ms/step - loss: 0.7898 - accuracy: 0.6087 - val_loss: 0.0000e+00 - val_accuracy: 0.0000e+00
Epoch 2/10
8/8 [==============================] - 0s 6ms/step - loss: 0.6710 - accuracy: 0.6500 - val_loss: 0.0000e+00 - val_accuracy: 0.0000e+00
Epoch 3/10
8/8 [==============================] - 0s 5ms/step - loss: 0.6748 - accuracy: 0.6500 - val_loss: 0.0000e+00 - val_accuracy: 0.0000e+00
Epoch 4/10
8/8 [==============================] - 0s 6ms/step - loss: 0.6716 - accuracy: 0.6370 - val_loss: 0.0000e+00 - val_accuracy: 0.0000e+00
Epoch 5/10
8/8 [==============================] - 0s 6ms/step - loss: 0.6085 - accuracy: 0.6326 - val_loss: 0.0000e+00 - val_accuracy: 0.0000e+00
Epoch 6/10
8/8 [==============================] - 0s 6ms/step - loss: 0.6744 - accuracy: 0.6326 - val_loss: 0.0000e+00 - val_accuracy: 0.0000e+00
Epoch 7/10
8/8 [==============================] - 0s 6ms/step - loss: 0.6102 - accuracy: 0.6522 - val_loss: 0.0000e+00 - val_accuracy: 0.0000e+00
Epoch 8/10
8/8 [==============================] - 0s 6ms/step - loss: 0.7032 - accuracy: 0.6109 - val_loss: 0.0000e+00 - val_accuracy: 0.0000e+00
Epoch 9/10
8/8 [==============================] - 0s 5ms/step - loss: 0.6283 - accuracy: 0.6717 - val_loss: 0.0000e+00 - val_accuracy: 0.0000e+00
Epoch 10/10
8/8 [==============================] - 0s 5ms/step - loss: 0.6120 - accuracy: 0.6652 - val_loss: 0.0000e+00 - val_accuracy: 0.0000e+00

So, definitely there is some issue with tensorflow implementation of fit.

I dug up the source, and it seems the part responsible for validation_data:

...
...
        # Run validation.
        if validation_data and self._should_eval(epoch, validation_freq):
          val_x, val_y, val_sample_weight = (
              data_adapter.unpack_x_y_sample_weight(validation_data))
          val_logs = self.evaluate(
              x=val_x,
              y=val_y,
              sample_weight=val_sample_weight,
              batch_size=validation_batch_size or batch_size,
              steps=validation_steps,
              callbacks=callbacks,
              max_queue_size=max_queue_size,
              workers=workers,
              use_multiprocessing=use_multiprocessing,
              return_dict=True)
          val_logs = {'val_' + name: val for name, val in val_logs.items()}
          epoch_logs.update(val_logs)

internally calls model.evaluate, as we have already established evaluate works fine, I realized the only culprit could be unpack_x_y_sample_weight.

So, I looked into the implementation:

def unpack_x_y_sample_weight(data):
  """Unpacks user-provided data tuple."""
  if not isinstance(data, tuple):
    return (data, None, None)
  elif len(data) == 1:
    return (data[0], None, None)
  elif len(data) == 2:
    return (data[0], data[1], None)
  elif len(data) == 3:
    return (data[0], data[1], data[2])

  raise ValueError("Data not understood.")

It's crazy, but if you just pass a tuple instead of a list, everything works fine due to the check inside unpack_x_y_sample_weight. (Your labels are missing after this step and somehow the data is getting fixed inside evaluate, so you're training with no reasonable labels, this seems like a bug but the documentation clearly states to pass tuple)

The following code gives correct validation accuracy and loss:

model.fit(X_train, y_train, validation_data=(X_train.to_numpy(), y_train.to_numpy()), 
epochs=10, batch_size=64)

Epoch 1/10
8/8 [==============================] - 0s 7ms/step - loss: 0.5832 - accuracy: 0.6696 - val_loss: 0.6892 - val_accuracy: 0.6674
Epoch 2/10
8/8 [==============================] - 0s 7ms/step - loss: 0.6385 - accuracy: 0.6804 - val_loss: 0.8984 - val_accuracy: 0.5565
Epoch 3/10
8/8 [==============================] - 0s 7ms/step - loss: 0.6822 - accuracy: 0.6391 - val_loss: 0.6556 - val_accuracy: 0.6739
Epoch 4/10
8/8 [==============================] - 0s 6ms/step - loss: 0.6276 - accuracy: 0.6609 - val_loss: 1.0691 - val_accuracy: 0.5630
Epoch 5/10
8/8 [==============================] - 0s 7ms/step - loss: 0.7048 - accuracy: 0.6239 - val_loss: 0.6474 - val_accuracy: 0.6326
Epoch 6/10
8/8 [==============================] - 0s 7ms/step - loss: 0.6545 - accuracy: 0.6500 - val_loss: 0.6659 - val_accuracy: 0.6043
Epoch 7/10
8/8 [==============================] - 0s 7ms/step - loss: 0.5796 - accuracy: 0.6913 - val_loss: 0.6891 - val_accuracy: 0.6435
Epoch 8/10
8/8 [==============================] - 0s 7ms/step - loss: 0.5915 - accuracy: 0.6891 - val_loss: 0.5307 - val_accuracy: 0.7152
Epoch 9/10
8/8 [==============================] - 0s 7ms/step - loss: 0.5571 - accuracy: 0.7000 - val_loss: 0.5465 - val_accuracy: 0.6957
Epoch 10/10
8/8 [==============================] - 0s 7ms/step - loss: 0.7133 - accuracy: 0.6283 - val_loss: 0.7046 - val_accuracy: 0.6413

So, as this seems to be a bug, I have just opened a relevant issue at Tensorflow Github repo:

https://github.com/tensorflow/tensorflow/issues/39370