# Keras Inference vs Training Results

**URL:** <https://discuss.ai.google.dev/t/keras-inference-vs-training-results/32329>\
**Category:** General Discussion\
**Tags:** custom-layer, keras, models\
**Created:** [September 14, 2022, 12:54pm UTC](https://discuss.ai.google.dev/t/keras-inference-vs-training-results/32329 "2022-09-14T12:54:24Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Kevin\_T](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/kevin_t/32/11983_2.png) [@Kevin\_T](https://discuss.ai.google.dev/u/Kevin_T)\
**Post date:** [September 14, 2022, 12:54pm UTC](https://discuss.ai.google.dev/t/keras-inference-vs-training-results/32329/1 "2022-09-14T12:54:24Z")

</div>

Hello, this is my first time here, so please correct me if I’m in the wrong forum.

I’m using a custom layer that scales the result of a Dense layer between 0 and 1:

```auto
class BoundingBoxNumber(tf.keras.layers.Layer):

	def __init__ (self, self_input_shape):
		super(BoundingBoxNumber,self). __init__ ()

		self.input_shape_custom = self_input_shape

		self.input_shape_accu = 1

		for item in self_input_shape:
			self.input_shape_accu *= item

		self.internal_dense_layer = tf.keras.layers.Dense(1, activation = 'tanh')

	@tf.function
	def call(self, inputs):

		inputs = tf.keras.layers.Flatten()(inputs)

		inputs.set_shape((1, self.input_shape_accu))

		output = self.internal_dense_layer(inputs)

		

		output = tf.divide(output, 2)

		output = tf.math.add(output, 0.5)
		

		return(output)

```

Then I’m using a custom training loop:

```auto
for epoch in range(50):

	print("Epoch:", epoch, "of 50")

	average_loss = 0

	for iter, item in enumerate(image):
		
		num_bounding_boxes = tf.shape(bb[iter])[0]

		float_target = tf.cast(1/num_bounding_boxes, tf.float32)

		with tf.GradientTape() as tape:

			logits = dense_bb_num_layer(item)
			
			loss = NumLoss(logits, float_target)

			print("Logits:", logits, "Target:", float_target, "Loss:", loss)
		

		average_loss += loss

		call_gradients = tape.gradient(loss, dense_bb_num_layer.trainable_weights)

		call_optimizer.apply_gradients(zip(call_gradients, dense_bb_num_layer.trainable_weights))

	average_loss /= len(image)

	print(average_loss)

```

and custom loss function:

```auto
def NumLoss(logits, expected):

	return((logits - expected)**2)

```

When running the layer in the custom training loop, the result is always exactly 0 or 1, whereas when it is outside (just calling BoundingBoxNumber()(input), I get a random float, as expected). My training data are all floats between 0 and 1, none of them are exactly zero or one.

---

<div class="post-metadata">

**Author:** ![Renu\_Patel](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/renu_patel/32/8957_2.png) [@Renu\_Patel](https://discuss.ai.google.dev/u/Renu_Patel)\
**Post date:** [January 18, 2024, 5:56pm UTC](https://discuss.ai.google.dev/t/keras-inference-vs-training-results/32329/3 "2024-01-18T17:56:48Z")

</div>

Hi @Kevin_T

Welcome to the TensorFlow Forum!

Please let us know if this issue still persists. If so, Could you please share the standalone code along with the dataset type and shape you are using for model training to better understand the issue? Thank you.