Update vLLM version support to include 0.14.0 and 0.14.1#5214
Merged
qgallouedec merged 2 commits intomainfrom Mar 4, 2026
Merged
Update vLLM version support to include 0.14.0 and 0.14.1#5214qgallouedec merged 2 commits intomainfrom
qgallouedec merged 2 commits intomainfrom
Conversation
qgallouedec
changed the title
qgallouedec
commented
Mar 3, 2026
Comment on lines
+539
to
+558
| def test_generate_with_params(self): | ||
| prompts = ["Hello, AI!", "Tell me a joke"] | ||
| completion_ids = self.client.generate(prompts, n=2, repetition_penalty=0.9, temperature=0.8, max_tokens=32)[ | ||
| "completion_ids" | ||
| ] | ||
|
|
||
| # Check that the output is a list | ||
| assert isinstance(completion_ids, list) | ||
|
|
||
| # Check that the number of generated sequences is 2 times the number of prompts | ||
| assert len(completion_ids) == 2 * len(prompts) | ||
|
|
||
| # Check that the generated sequences are lists of integers | ||
| for seq in completion_ids: | ||
| assert all(isinstance(tok, int) for tok in seq) | ||
|
|
||
| # Check that the length of the generated sequences is less than or equal to 32 | ||
| for seq in completion_ids: | ||
| assert len(seq) <= 32 | ||
|
|
Member
Author
There was a problem hiding this comment.
not specific to vllm 0.14, but I realized that this test case was missing
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 59252637ef
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
qgallouedec
commented
Mar 3, 2026
Comment on lines
+668
to
+687
| def test_generate_with_params(self): | ||
| prompts = ["Hello, AI!", "Tell me a joke"] | ||
| completion_ids = self.client.generate(prompts, n=2, repetition_penalty=0.9, temperature=0.8, max_tokens=32)[ | ||
| "completion_ids" | ||
| ] | ||
|
|
||
| # Check that the output is a list | ||
| assert isinstance(completion_ids, list) | ||
|
|
||
| # Check that the number of generated sequences is 2 times the number of prompts | ||
| assert len(completion_ids) == 2 * len(prompts) | ||
|
|
||
| # Check that the generated sequences are lists of integers | ||
| for seq in completion_ids: | ||
| assert all(isinstance(tok, int) for tok in seq) | ||
|
|
||
| # Check that the length of the generated sequences is less than or equal to 32 | ||
| for seq in completion_ids: | ||
| assert len(seq) <= 32 | ||
|
|
Member
Author
There was a problem hiding this comment.
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
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.
Summary
Extend TRL’s vLLM support to
0.14.0and0.14.1.Changes
vLLM 0.14.0 introduced a breaking change: DP for dense models now errors out. From vllm-project/vllm#30739.
Reproducer and traceback
In my understanding, they say that scaling DP for dense models is always detrimental to performance. Which is surprising considering my old benchmark. Anyways, I recommend aligning with vLLM recommendations, and discourage scaling DP for dense model when even possible (vllm<0.14).
Tests