Fix take_along_axis with TensorFlow with non-int32 indices.

[hertschuh]

Change #21239 broke one use case when the axis dimension is dynamic, the type of the indices is not int32, and the op is run in graph mode.

Note that the additional unit tests don't actually cover this.

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 82.59%. Comparing base (6ddaefb) to head (8751fdb).

Additional details and impacted files

@@           Coverage Diff           @@
##           master   #21256   +/-   ##
=======================================
  Coverage   82.59%   82.59%           
=======================================
  Files         564      564           
  Lines       54594    54594           
  Branches     8483     8483           
=======================================
  Hits        45092    45092           
  Misses       7415     7415           
  Partials     2087     2087           

Flag	Coverage Δ
keras	82.40% <100.00%> (ø)
keras-jax	63.66% <0.00%> (ø)
keras-numpy	58.80% <0.00%> (ø)
keras-openvino	32.99% <0.00%> (ø)
keras-tensorflow	64.09% <100.00%> (ø)
keras-torch	63.74% <0.00%> (ø)

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.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[fchollet]

Thanks for the fix!