[Bugfix][Frontend] Preserve abort finish_reason for scale-out token streams

[Sunt-ing]

Purpose

In token mode, /inference/v1/generate streams skip engine outputs that have no new token IDs. An abort always ends with such an output (token_ids=[], finish_reason="abort"), so an aborted stream ends with [DONE] and no finish_reason. The only exception is when the abort happens to be merged into a pending token delta. Text mode (#58588) and the Rust frontend already emit every terminal output.

• Every output that has a finish_reason is now emitted, in both output modes.
• The per-choice token counter is now sized by sampling_params.n. It was sized from the first output, which with n > 1 may hold only choice 0. The next choice then raised IndexError, and the stream ended in a 500 error chunk.

Test Plan

Live server on H200 with Qwen/Qwen3-0.6B. The same requests went to main (bc21cba) and to this PR:

vllm serve Qwen/Qwen3-0.6B --max-model-len 2048 --no-enable-prefix-caching --tokens-only
vllm serve Qwen/Qwen3-0.6B --max-model-len 2048 --no-enable-prefix-caching --enable-scale-out --max-num-seqs 1

Each request was a streamed /inference/v1/generate call with max_tokens: 900, ignore_eos: true and include_usage. It was aborted through /abort_requests (or /inference/v1/abort_requests) once the first token chunk arrived. With --max-num-seqs 1, a second request could also be aborted while still queued.

Test Result

Case	main	this PR
tokens mode, abort mid-stream (also with logprobs / flat_logprobs)	no finish_reason unless the abort merged into a token chunk	final chunk token_ids: [], finish_reason: "abort" (logprobs: {"content": []})
abort while queued (0 tokens)	no choice chunk	one abort chunk
n=2, abort mid-stream	list index out of range 500 chunk in 4/7 runs	both choices get abort, 12/12 runs
text-mode abort, non-stream abort, length / EOS stop	—	unchanged

Usage completion_tokens matched the streamed token count in every case.

[claude]

Claude Code Review

This pull request is from a fork — automated review is disabled. A repository maintainer can comment @claude review to run a one-time review.

[Sunt-ing]

@DarkLight1337 A simple PR. PTAL. Thanks~

[mergify]

This pull request has merge conflicts that must be resolved before it can be
merged. Please rebase the PR, @Sunt-ing.

https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/working-with-forks/syncing-a-fork

[Sunt-ing]

Hi @aoshen02, could you please take a look at this simple bugfix? Thanks~

[mergify]

This pull request has merge conflicts that must be resolved before it can be
merged. Please rebase the PR, @Sunt-ing.

https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/working-with-forks/syncing-a-fork

[sfeng33]

/ci run --allow-stale

[github-actions]

✅ Triggered Buildkite CI #92775 for commit 1dcd0c41d357.

⚠️ This PR is 6 commits behind upstream main. Running CI at your own risk because --allow-stale was requested; outdated CI configuration may cause failures. Before merging, merge or rebase onto the latest main, then rerun /ci run on the latest PR commit.

[vllm-agent]

CI selector (shadow): 2 test steps (5 jobs) instead of 35 (48 jobs)

Shadow mode: this changes nothing about what CI runs. It shows what the evidence-based selector would pick for this PR, next to today's rules. How it works.

Feedback welcome: reply here if it would skip a step this change needs, or runs something unrelated.

steps (jobs)	Today's rules	Selector	Would skip	Would add
NVIDIA, CPU and others	35 (48)	2 (5)	33 (43)	0 (0)
AMD mirrors	34 (42)	2 (5)	32 (37)	0 (0)

Selector would run (2)

• entrypoints-integration-api-server ×4
• rust-frontend-serve-admin-coverage

Would skip (today's rules run them) (33)

• basic-correctness ×2
• basic-correctness-cpu-offload
• basic-correctness-cumem
• basic-correctness-prefetch-offload
• basic-correctness-sleep-mode
• basic-models-test-other-cpu
• basic-models-tests-initialization
• basic-models-tests-other
• benchmarks-cli-test
• cpu-language-generation-and-pooling-model-tests ×3
• entrypoints-integration-api-server-generate
• entrypoints-integration-api-server-openai-chat_completion
• entrypoints-integration-api-server-openai-completion
• entrypoints-integration-llm
• entrypoints-integration-multimodal
• entrypoints-integration-pooling
• entrypoints-integration-responses-api
• entrypoints-integration-speech_to_text
• entrypoints-unit-tests
• examples
• kernels-fla-ops-test-b200
• kernels-mhc-test-b200
• kernels-root-misc-test-b200
• language-models-tests-granite-l4-compatibility
• language-models-tests-hybrid ×2
• language-models-tests-standard
• multi-modal-models-standard-1-qwen2
• multi-modal-models-standard-2-qwen3-gemma
• multi-modal-models-standard-3-llava-qwen2-vl
• multi-modal-models-standard-4-other-whisper
• multi-modal-processor ×4
• multi-modal-processor-cpu ×4
• scale-out-ec-e2e-2-gpus

Would add (today's rules do not run them) (0)

none

AMD mirrors: would skip (32)

• basic-correctness ×2
• basic-correctness-cpu-offload
• basic-correctness-cumem
• basic-correctness-prefetch-offload
• basic-correctness-sleep-mode
• basic-models-tests-initialization
• basic-models-tests-other
• benchmarks-cli-test
• entrypoints-integration-api-server-generate
• entrypoints-integration-api-server-openai-chat_completion
• entrypoints-integration-api-server-openai-completion
• entrypoints-integration-llm
• entrypoints-integration-multimodal
• entrypoints-integration-pooling
• entrypoints-integration-responses-api
• entrypoints-integration-speech_to_text
• entrypoints-unit-tests
• examples
• kernels-fla-ops-test-b200
• kernels-mhc-test-b200
• kernels-root-misc-test-b200
• language-models-tests-granite-l4-compatibility
• language-models-tests-hybrid ×2
• language-models-tests-standard
• multi-modal-models-standard-1-qwen2
• multi-modal-models-standard-2-qwen3-gemma
• multi-modal-models-standard-3-llava-qwen2-vl
• multi-modal-models-standard-4-other-whisper
• multi-modal-processor ×4
• platform-tests
• pytorch-compilation-passes-unit-tests
• scale-out-ec-e2e-2-gpus

AMD mirrors: would add (0)

none

_{3 changed files · base bc21cba967 · head 1dcd0c41d3 · Python record: build 92769 at 84bcbc6264 · kernel record: table 84bcbc6 (build 92769), map 84bcbc6 · not counted: 10 build steps, 5 A100 steps the generator no longer emits, 0 optional steps the selector would also run}