Fix issue with Masking layer with Tensor as mask_value

[Surya2k1]

Currently , the Masking layer accepting Tensor also for argument mask_value . After saving and loading the model the mask_value no longer a tensor. Hence the issue. The issue with be an model which have Masking layer and uses a Tensor as mask_value.

Either the Masking layer should not allow Tensors for mask_value or while reloading the model we need to take care of this explicitly.

I am proposing to handle mask_value explicitly during reloading the model.

Might fix #20706

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 82.01%. Comparing base (5df8fb9) to head (3428a6a).

Additional details and impacted files

@@           Coverage Diff           @@
##           master   #20791   +/-   ##
=======================================
  Coverage   82.01%   82.01%           
=======================================
  Files         557      557           
  Lines       52014    52017    +3     
  Branches     8037     8038    +1     
=======================================
+ Hits        42657    42662    +5     
+ Misses       7403     7402    -1     
+ Partials     1954     1953    -1     

Flag	Coverage Δ
keras	81.83% <100.00%> (+<0.01%)	⬆️
keras-jax	64.23% <100.00%> (+0.01%)	⬆️
keras-numpy	58.95% <33.33%> (-0.01%)	⬇️
keras-openvino	29.89% <33.33%> (+<0.01%)	⬆️
keras-tensorflow	64.79% <100.00%> (+<0.01%)	⬆️
keras-torch	64.16% <100.00%> (-0.02%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[fchollet]

Thanks for the PR!

You can fix the code format by running sh shell/format.sh

[fchollet]

You can actually remove this line! deserialize_keras_object should work in all circumstances.

[Surya2k1]

You can actually remove this line! deserialize_keras_object should work in all circumstances.

No. During model building with Tensor it will fail. For example consider below example code where we found the issue. Without this commit the below code is fine while building but fails when saved and reloaded.

model = keras.models.Sequential()
model.add(keras.layers.Masking(mask_value=keras.ops.convert_to_tensor([0.0])))

This will fail with Type Error. Hence I put that check to ensure only serialized tensor will go for deserialization and not other inputs and also not any random dict.

TypeError: Could not parse config: [0.]

    model.add(keras.layers.Masking(mask_value=keras.ops.convert_to_tensor(np.array([0.0]))))
              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "D:\OSS\Keras\keras\keras\src\layers\core\masking.py", line 51, in __init__
    mask_value = deserialize_keras_object(mask_value)
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "D:\OSS\Keras\keras\keras\src\saving\serialization_lib.py", line 591, in deserialize_keras_object
    raise TypeError(f"Could not parse config: {config}")
TypeError: Could not parse config: [0.]