fix(deepseek-v41): backport upstream Responses text-part support

[voipmonitor]

Purpose

Accept valid Responses API text parts in the DeepSeek V4.1 tokenizer.
input_text and output_text render identically to the existing Chat
Completions text form. Image handling, separators, model execution and
sampling are unchanged.

Backport of merged upstream vllm-project/vllm#56299,
which fixes #56297.
CedricHuang's authorship and original commit attribution are preserved.
Only its focused parity test is imported; unrelated upstream reasoning-budget
fixtures are not substituted for LIL's reference-encoder tests.

No equivalent open LIL fix was found. #787 addresses tool namespaces and
reminders, not Responses text-part types. Both changes compose in beta.

Validation

Status: qualified within the following tokenizer and live API checks.

• Composed tool serialization, V4/V4.1 tokenizer and DeepSeek parser suites:
  187 passed, including both Responses parity cases.
• Both user input_text and assistant output_text produce the same prompt
  as text; input messages remain unchanged.
• Applicable pre-commit hooks pass.
• Six HTTP cases pass on DS4.1 TP4/DSpark K7 at temperature 1 in a built
  image without source mounts: string input, structured input_text, streamed
  completion, assistant output_text history, synthetic tool-call/result
  history, and an image request with detail: auto.
• No kernel or model-quality improvement is claimed.

Raw HTTP checks, native results and source identities.

AI-assisted backport and validation; human maintainer review remains required.
Tracked with the source review checklist in #773. Integration is restricted to
integration/beta; this PR does not authorize a JJ merge.

[coderabbitai]

Warning

Review limit reached

Next included review available in 47 minutes.

Check out review usage here.

View limit details

Limit details: You’ve used the included review currently available.

You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository.

Learn how review limits work.

Review configuration:

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Advanced

Run ID: c4f03e38-c42e-4d33-86ea-f8ce1d6dd4d0

📥 Commits

Reviewing files that changed from the base of the PR and between 8e1f1e5 and 5ee159b.

📒 Files selected for processing (2)

• tests/tokenizers_/test_deepseek_v41.py
• vllm/tokenizers/deepseek_v41.py

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[voipmonitor]

@coderabbitai review

[coderabbitai]

⚠️ Action not completed

Review rate limited.

Note: CodeRabbit is an incremental review system and does not re-review already reviewed commits. This command is applicable only when automatic reviews are paused.