Add edge case tests for out-of-range token id decoding in Qwen2 tokenizer

[saslifat-gif]

The Qwen2 tokenizer test file had no custom test methods — only integration
constants inherited from TokenizerTesterMixin.

This PR adds a test documenting two untested edge cases in decode():

Before (no test, behavior undocumented):

tok.decode([999999])  # silently returns ''
tok.decode([-1])      # raises: out of range integral type conversion attempted

After (behavior documented and regression-protected):

• Token ids beyond len(tokenizer) (e.g. 999999) silently return an empty
  string — tested with assertEqual(decoded, "")
• Negative token ids raise an exception — tested with assertRaises(Exception)

Both behaviors verified consistent between Qwen2Tokenizer (slow) and fast
tokenizer on Qwen/Qwen2.5-VL-7B-Instruct.

[github-actions]

[For maintainers] Suggested jobs to run (before merge)

run-slow: qwen2

[github-actions]

View the CircleCI Test Summary for this PR:

https://huggingface.co/spaces/transformers-community/circle-ci-viz?pr=45191&sha=81e61e

[Rocketknight1]

Hi @saslifat-gif, we'd prefer no small PRs right now that aren't significant bugfixes or feature additions! We're trying to cut down on the overall volume because of agent spam.