Skip to content

[tests] Unskip DeBERTaV2 tokenizer parity tests; re-enable fast/slow checks#40473

Open
asifh0ssain wants to merge 2 commits intohuggingface:mainfrom
asifh0ssain:fix/deberta-v2-tokenizer-parity
Open

[tests] Unskip DeBERTaV2 tokenizer parity tests; re-enable fast/slow checks#40473
asifh0ssain wants to merge 2 commits intohuggingface:mainfrom
asifh0ssain:fix/deberta-v2-tokenizer-parity

Conversation

@asifh0ssain
Copy link
Copy Markdown

This PR re-enables two previously skipped parity tests for the DeBERTaV2 tokenizer:

  • test_sentencepiece_tokenize_and_convert_tokens_to_string
  • test_sentencepiece_tokenize_and_decode

Both tests now call the mixin implementations to verify fast and slow tokenizers produce consistent results.

✅ No API changes.
✅ Code passes local lint/format checks.
Hugging Face CI will confirm full suite.

@github-actions
Copy link
Copy Markdown
Contributor

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

run-slow: deberta_v2

@Rocketknight1
Copy link
Copy Markdown
Member

The tests appear to be failing in this PR 😅

@asifh0ssain
Copy link
Copy Markdown
Author

The tests appear to be failing in this PR 😅

Working on it. Thanks for checking!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants