Merge pull request #17746 from Frightera:frightera_small_loss_fixes

PiperOrigin-RevId: 522400412

Author: tensorflower-gardener

keras/backend.py

@@ -5567,8 +5567,12 @@ def categorical_crossentropy(target, output, from_logits=False, axis=-1):
             labels=target, logits=output, axis=axis
         )
 
-    # scale preds so that the class probas of each sample sum to 1
+    # Adjust the predictions so that the probability of
+    # each class for every sample adds up to 1
+    # This is needed to ensure that the cross entropy is
+    # computed correctly.
     output = output / tf.reduce_sum(output, axis, True)
+
     # Compute cross entropy from probabilities.
     epsilon_ = _constant_to_tensor(epsilon(), output.dtype.base_dtype)
     output = tf.clip_by_value(output, epsilon_, 1.0 - epsilon_)
@@ -5845,28 +5849,29 @@ def binary_focal_crossentropy(
     where `alpha` is a float in the range of `[0, 1]`.
 
     Args:
-      target: A tensor with the same shape as `output`.
-      output: A tensor.
-      apply_class_balancing: A bool, whether to apply weight balancing on the
-        binary classes 0 and 1.
-      alpha: A weight balancing factor for class 1, default is `0.25` as
-        mentioned in the reference. The weight for class 0 is `1.0 - alpha`.
-      gamma: A focusing parameter, default is `2.0` as mentioned in the
-        reference.
-      from_logits: Whether `output` is expected to be a logits tensor. By
-        default, we consider that `output` encodes a probability distribution.
+        target: A tensor with the same shape as `output`.
+        output: A tensor.
+        apply_class_balancing: A bool, whether to apply weight balancing on the
+            binary classes 0 and 1.
+        alpha: A weight balancing factor for class 1, default is `0.25` as
+            mentioned in the reference. The weight for class 0 is `1.0 - alpha`.
+        gamma: A focusing parameter, default is `2.0` as mentioned in the
+            reference.
+        from_logits: Whether `output` is expected to be a logits tensor. By
+            default, we consider that `output` encodes a probability
+            distribution.
 
     Returns:
-      A tensor.
+        A tensor.
     """
-    sigmoidal = tf.__internal__.smart_cond.smart_cond(
-        from_logits,
-        lambda: sigmoid(output),
-        lambda: output,
-    )
+
+    sigmoidal = sigmoid(output) if from_logits else output
+
     p_t = target * sigmoidal + (1 - target) * (1 - sigmoidal)
+
     # Calculate focal factor
     focal_factor = tf.pow(1.0 - p_t, gamma)
+
     # Binary crossentropy
     bce = binary_crossentropy(
         target=target,
@@ -5894,7 +5899,7 @@ def sigmoid(x):
     Returns:
         A tensor.
     """
-    return tf.sigmoid(x)
+    return tf.math.sigmoid(x)
 
 
 @keras_export("keras.backend.hard_sigmoid")