fix: remove duplicate failing raggedtensor test when you have a GPU on the system

[ikrommyd]

tests/test_3210_to_raggedtensor_from_raggedtensor.py::test_convert_from_raggedtensor is failing for me locally when I have a GPU on my system because tensorflow automatically uses the GPU for all operations as the default.

(ak-uproot) ➜  awkward git:(main) ✗ pytest tests/test_3210_to_raggedtensor_from_raggedtensor.py::test_convert_from_raggedtensor
================================================================================================================================ test session starts ================================================================================================================================
platform linux -- Python 3.12.10, pytest-8.3.5, pluggy-1.5.0
rootdir: /home/iason/Dropbox/work/coffea_dev/awkward
configfile: pyproject.toml
plugins: cov-6.1.1, anyio-4.9.0, xdist-3.6.1
collected 1 item

tests/test_3210_to_raggedtensor_from_raggedtensor.py F                                                                                                                                                                                                                        [100%]

===================================================================================================================================== FAILURES ======================================================================================================================================
__________________________________________________________________________________________________________________________ test_convert_from_raggedtensor ___________________________________________________________________________________________________________________________

    def test_convert_from_raggedtensor():
        tf_array1 = tf.RaggedTensor.from_row_splits(
            values=[1.1, 2.2, 3.3, 4.4, 5.5], row_splits=[0, 2, 3, 3, 5]
        )

        ak_array1 = ak.contents.ListOffsetArray(offsets1, content1)
        result1 = ak.to_layout(from_raggedtensor(tf_array1), allow_record=False)
        assert (
            result1.content.data == ak.to_backend(np_array1, result1.backend).layout.data
        ).all()
>       assert (result1.offsets.data == [0, 2, 3, 3, 5]).all()

ak_array1  = <ListOffsetArray len='4'>
    <offsets><Index dtype='int64' len='5'>[0 2 3 3 5]</Index></offsets>
    <content><NumpyArray dtype='float32' len='5'>[1.1 2.2 3.3 4.4 5.5]</NumpyArray></content>
</ListOffsetArray>
result1    = <ListOffsetArray len='4'>
    <offsets><Index dtype='int64' len='5'>[0 2 3 3 5]</Index></offsets>
    <content><NumpyArray dtype='float32' len='5'>[1.1 2.2 3.3 4.4 5.5]</NumpyArray></content>
</ListOffsetArray>
tf_array1  = <tf.RaggedTensor [[1.1, 2.2], [3.3], [], [4.4, 5.5]]>

tests/test_3210_to_raggedtensor_from_raggedtensor.py:113:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

>   ???
E   TypeError: operand type(s) all returned NotImplemented from __array_ufunc__(<ufunc 'equal'>, '__call__', array([0, 2, 3, 3, 5]), [0, 2, 3, 3, 5]): 'ndarray', 'list'


cupy/_core/core.pyx:1264: TypeError
------------------------------------------------------------------------------------------------------------------------------- Captured stderr call --------------------------------------------------------------------------------------------------------------------------------
I0000 00:00:1744507395.989245  657278 gpu_device.cc:2019] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 12937 MB memory:  -> device: 0, name: NVIDIA GeForce RTX 4080 SUPER, pci bus id: 0000:01:00.0, compute capability: 8.9
============================================================================================================================== short test summary info ==============================================================================================================================
FAILED tests/test_3210_to_raggedtensor_from_raggedtensor.py::test_convert_from_raggedtensor - TypeError: operand type(s) all returned NotImplemented from __array_ufunc__(<ufunc 'equal'>, '__call__', array([0, 2, 3, 3, 5]), [0, 2, 3, 3, 5]): 'ndarray', 'list'
================================================================================================================================= 1 failed in 2.29s =================================================================================================================================

I can make it succeed if I specify CUDA_VISIBLE_DEVICES='' pytest tests/test_3210_to_raggedtensor_from_raggedtensor.py::test_convert_from_raggedtensor so that tensorflow doesn't see the GPU.
Well this test is actually a duplicate and the line below it tests the same thing properly! I'm just removing the bad line here. It's good to take care of such things for when a GPU ci is available.

[ianna]

@ikrommyd - thanks for fixing it!