Mark prompt logprobs as incompatible with prompt embeds at API level #25077
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…evel Signed-off-by: Andrew Sansom <[email protected]>
qthequartermasterman
requested review from
DarkLight1337,
NickLucche,
aarnphm,
alexm-redhat,
chaunceyjiang,
comaniac,
njhill,
robertgshaw2-redhat,
simon-mo,
youkaichao and
zhuohan123
as code owners
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request introduces a validation to prohibit the simultaneous use of prompt_logprobs and prompt_embeds in completion requests, addressing an incompatibility. The validation is implemented at two levels: at the API entrypoint in serving_completion.py to provide a clear error to the user, and at the engine level in llm_engine.py as a robust safeguard. A corresponding unit test has been added to verify that the API correctly rejects such requests. The changes are logical, well-tested, and effectively prevent the unsupported combination of parameters.
DarkLight1337
approved these changes
Sep 17, 2025
github-actions
bot
added
the
ready
5 tasks
qthequartermasterman
added a commit
to protopia-ai/vllm
that referenced
this pull request
Sep 17, 2025
Signed-off-by: Andrew Sansom <[email protected]>
|
@DarkLight1337 is it normal for Readthedocs to take over 10 hours to complete? |
|
No, let me rebuild it |
845473182
pushed a commit
to dsxsteven/vllm_splitPR
that referenced
this pull request
Sep 18, 2025
…litPR into model_register * 'model_register' of https://github.com/dsxsteven/vllm_splitPR: (138 commits) Retrieve `sliding_window` from text config in Gemma3 MM (vllm-project#25085) [Docs] Fix API Reference (vllm-project#25140) [Kernel] Better inf handling for grouped topk cu (vllm-project#24886) [CLI] Use streaming in CLI chat and completion commands (vllm-project#23769) [benchmark] add peak throughput metrics and plot (vllm-project#23867) [Spec Decode] Efficient padded speculation (vllm-project#24539) [V0 Deprecation] Remove more V0 tests (vllm-project#25117) [EPLB] Add EPLB support for hunyuan_v1 (vllm-project#23078) [XPU] Whisper model support on XPU Platform (vllm-project#25123) Mark prompt logprobs as incompatible with prompt embeds at API level (vllm-project#25077) [Model] enable data parallel for InternVL vision encoder (vllm-project#23909) [Kernels] Overlap shared experts with combine instead of dispatch (vllm-project#24254) [Bugfix][Qwen3-Next] add prefixes to shared_expert in qwen3-next and mlp in qwen2moe to successfully load ignored params in quantized models (vllm-project#24960) [Core][MM] Cleanup `MultiModalCache` (vllm-project#25006) [Docs] Clean up the contributing README (vllm-project#25099) [MM Encoder] Apply DP ViT for Qwen3-VL model series (vllm-project#24955) [Kernels] Enable DeepGEMM by default (vllm-project#24462) [V0 Deprecation] Skip PP test (vllm-project#25128) [V0 Deprecation] Remove misc V0 tests (vllm-project#25118) [V0 Deprecation] Remove V0 Tracing & Metrics tests (vllm-project#25115) ...
debroy-rh
pushed a commit
to debroy-rh/vllm
that referenced
this pull request
Sep 19, 2025
…llm-project#25077) Signed-off-by: Andrew Sansom <[email protected]>
FeiDaLI
pushed a commit
to FeiDaLI/vllm
that referenced
this pull request
Sep 25, 2025
…llm-project#25077) Signed-off-by: Andrew Sansom <[email protected]>
charlifu
pushed a commit
to ROCm/vllm
that referenced
this pull request
Sep 25, 2025
…llm-project#25077) Signed-off-by: Andrew Sansom <[email protected]> Signed-off-by: charlifu <[email protected]>
xuebwang-amd
pushed a commit
to xuebwang-amd/vllm
that referenced
this pull request
Oct 10, 2025
…llm-project#25077) Signed-off-by: Andrew Sansom <[email protected]> Signed-off-by: xuebwang-amd <[email protected]>
choprahetarth
pushed a commit
to Tandemn-Labs/vllm
that referenced
this pull request
Oct 11, 2025
…llm-project#25077) Signed-off-by: Andrew Sansom <[email protected]>
xuebwang-amd
pushed a commit
to xuebwang-amd/vllm
that referenced
this pull request
Oct 24, 2025
…llm-project#25077) Signed-off-by: Andrew Sansom <[email protected]> Signed-off-by: xuebwang-amd <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Purpose
The implementation of Prompt Logprobs is incompatible with Prompt Embeds in both the existing v0 engine and in #24278. In that PR is was requested by @DarkLight1337 to block requests with both prompt embeds and prompt logprobs at the API level.
This PR has changes originally in #24278, but since these fixes apply to both engines, and aren't strictly related to enabling prompt embedding support in the v1 engine, I figured it would be a good idea to split it into a separate PR for easier review and to reduce the diff in #24278.
Test Plan
Updated unit test with prompt embeds to explicitly ban requests that have both prompt embeds and prompt logprobs.
Test Result
New tests pass locally. Pending CI, it should be good to go. The full CI of the original PR #24278 never failed related to these changes.
Essential Elements of an Effective PR Description Checklist
supported_models.mdandexamplesfor a new model.