Fix: Return Attention Scores when return_attention_scores=True (#20684)

* Fix: Ensure Attention Layer Returns Attention Scores when `return_attention_scores=True`

This pull request addresses an issue in the Attention layer where the return_attention_scores parameter wasn't correctly handled in the compute_output_shape method. This fix ensures that attention scores are returned when return_attention_scores=True.

## Changes Made
Modified compute_output_shape method to return the shape of both the attention output and the attention scores when return_attention_scores=True.

* Formatting

* Fixed score return and added unit tests for return_attention_scores=True

* Removed debug print statement

Author: Furkan-rgb

keras/src/layers/attention/attention.py

@@ -1,6 +1,7 @@
 from keras.src import backend
 from keras.src import ops
 from keras.src.api_export import keras_export
+from keras.src.backend import KerasTensor
 from keras.src.layers.layer import Layer
 
 
@@ -84,6 +85,8 @@ def __init__(
                 f"Received: score_mode={score_mode}"
             )
 
+        self._return_attention_scores = False
+
     def build(self, input_shape):
         self._validate_inputs(input_shape)
         self.scale = None
@@ -217,6 +220,7 @@ def call(
         use_causal_mask=False,
     ):
         self._validate_inputs(inputs=inputs, mask=mask)
+        self._return_attention_scores = return_attention_scores
         q = inputs[0]
         v = inputs[1]
         k = inputs[2] if len(inputs) > 2 else v
@@ -226,16 +230,17 @@ def call(
         scores_mask = self._calculate_score_mask(
             scores, v_mask, use_causal_mask
         )
-        result, attention_scores = self._apply_scores(
+        attention_output, attention_scores = self._apply_scores(
             scores=scores, value=v, scores_mask=scores_mask, training=training
         )
         if q_mask is not None:
             # Mask of shape [batch_size, Tq, 1].
             q_mask = ops.expand_dims(q_mask, axis=-1)
-            result *= ops.cast(q_mask, dtype=result.dtype)
+            attention_output *= ops.cast(q_mask, dtype=attention_output.dtype)
         if return_attention_scores:
-            return result, attention_scores
-        return result
+            return (attention_output, attention_scores)
+        else:
+            return attention_output
 
     def compute_mask(self, inputs, mask=None):
         self._validate_inputs(inputs=inputs, mask=mask)
@@ -244,8 +249,49 @@ def compute_mask(self, inputs, mask=None):
         return ops.convert_to_tensor(mask[0])
 
     def compute_output_shape(self, input_shape):
-        """Returns shape of value tensor dim, but for query tensor length"""
-        return (*input_shape[0][:-1], input_shape[1][-1])
+        query_shape, value_shape, key_shape = input_shape
+        if key_shape is None:
+            key_shape = value_shape
+
+        output_shape = (*query_shape[:-1], value_shape[-1])
+        if self._return_attention_scores:
+            scores_shape = (query_shape[0], query_shape[1], key_shape[1])
+            return output_shape, scores_shape
+        return output_shape
+
+    def compute_output_spec(
+        self,
+        inputs,
+        mask=None,
+        return_attention_scores=False,
+        training=None,
+        use_causal_mask=False,
+    ):
+        # Validate and unpack inputs
+        self._validate_inputs(inputs, mask)
+        query = inputs[0]
+        value = inputs[1]
+        key = inputs[2] if len(inputs) > 2 else value
+
+        # Compute primary output shape
+        output_shape = self.compute_output_shape(
+            [query.shape, value.shape, key.shape]
+        )
+        output_spec = KerasTensor(output_shape, dtype=self.compute_dtype)
+
+        # Handle attention scores if requested
+        if self._return_attention_scores:
+            scores_shape = (
+                query.shape[0],
+                query.shape[1],
+                key.shape[1],
+            )  # (batch_size, Tq, Tv)
+            attention_scores_spec = KerasTensor(
+                scores_shape, dtype=self.compute_dtype
+            )
+            return (output_spec, attention_scores_spec)
+
+        return output_spec
 
     def _validate_inputs(self, inputs, mask=None):
         """Validates arguments of the call method."""

keras/src/layers/attention/attention_test.py

@@ -358,3 +358,62 @@ def test_attention_compute_output_shape(self):
             ),
             output.shape,
         )
+
+    def test_return_attention_scores_true(self):
+        """Test that the layer returns attention scores along with outputs."""
+        # Generate dummy input data
+        query = np.random.random((2, 8, 16)).astype(np.float32)
+        value = np.random.random((2, 4, 16)).astype(np.float32)
+
+        # Initialize the Attention layer
+        layer = layers.Attention()
+
+        # Call the layer with return_attention_scores=True
+        output, attention_scores = layer(
+            [query, value], return_attention_scores=True
+        )
+
+        # Check the shape of the outputs
+        self.assertEqual(output.shape, (2, 8, 16))  # Output shape
+        self.assertEqual(
+            attention_scores.shape, (2, 8, 4)
+        )  # Attention scores shape
+
+    def test_return_attention_scores_true_and_tuple(self):
+        """Test that the layer outputs are a tuple when
+        return_attention_scores=True."""
+        # Generate dummy input data
+        query = np.random.random((2, 8, 16)).astype(np.float32)
+        value = np.random.random((2, 4, 16)).astype(np.float32)
+
+        # Initialize the Attention layer
+        layer = layers.Attention()
+
+        # Call the layer with return_attention_scores=True
+        outputs = layer([query, value], return_attention_scores=True)
+
+        # Check that outputs is a tuple
+        self.assertIsInstance(
+            outputs, tuple, "Expected the outputs to be a tuple"
+        )
+
+    def test_return_attention_scores_true_tuple_then_unpack(self):
+        """Test that outputs can be unpacked correctly."""
+        # Generate dummy input data
+        query = np.random.random((2, 8, 16)).astype(np.float32)
+        value = np.random.random((2, 4, 16)).astype(np.float32)
+
+        # Initialize the Attention layer
+        layer = layers.Attention()
+
+        # Call the layer with return_attention_scores=True
+        outputs = layer([query, value], return_attention_scores=True)
+
+        # Unpack the outputs
+        output, attention_scores = outputs
+
+        # Check the shape of the unpacked outputs
+        self.assertEqual(output.shape, (2, 8, 16))  # Output shape
+        self.assertEqual(
+            attention_scores.shape, (2, 8, 4)
+        )  # Attention scores shape