fix(testing): Fix LayoutXLM tokenization test and LightOnOCR SDPA flash test failures on main CI

[harshaljanjani]

What does this PR do?

The following failing tests were identified and fixed in this PR:

→ LayoutXLM: This PR (rm slow tokenizers) changed models/auto/tokenization_auto.py to map configs to single tokenizer classes instead of (slow, fast) tuples. It did touch tests/test_tokenization_common.py; having said that, it didn't update the merge_model_tokenizer_mappings fn so the [0] and [1] subscripts fail now.
→ LightOnOCR: This commit (Add LightOnOCR model implementation) added the LightOnOCR model which internally uses Pixtral's vision encoder. The skip list for test_sdpa_can_dispatch_on_flash already included "pixtral" (added on Jun 27, 2025), but LightOnOCR was not added to the same list despite having identical behavior (internally generates a block attention mask incompatible with Flash Attention), this should fix that :)

cc: @Rocketknight1

→ After: Both tests correctly skip gracefully (ref screenshots). merge_model_tokenizer_mappings should handle V5's single-class tokenizer mapping, and lighton_ocr should be skipped in test_sdpa_can_dispatch_on_flash alongside pixtral!

CI Failures:

Before the fix (feel free to cross-check; these errors are reproducible):

After the fix (feel free to cross-check):

Before submitting

• This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• Did you read the contributor guideline,
  Pull Request section?
• Was this discussed/approved via a Github issue or the forum? Please add a link
  to it if that's the case.
• Did you make sure to update the documentation with your changes? Here are the
  documentation guidelines, and
  here are tips on formatting docstrings.
• Did you fix any necessary existing tests?