[Model] GPT2ForSequenceClassification model

[nie3e]

Support GPT2ForSequenceClassification model.
Tested on nie3e/sentiment-polish-gpt2-small

Test:

curl -X 'POST' \
  'http://127.0.0.1:8000/classify' \
  -H 'Content-Type: application/json' \
  -d '{
  "model": "nie3e/sentiment-polish-gpt2-small",
  "input": ["Przestronny hotel , jasny , z dużymi oknami .", "Położony całkiem blisko centrum ."]
}'

Result:

{"id":"classify-6619cecdb01a4bf8900df136a9b33b15","object":"list","created":1749994841,"model":"nie3e/sentiment-polish-gpt2-small","data":[{"index":0,"label":"POSITIVE","probs":[6.198883056640625e-06,1.7881393432617188e-07,1.0,7.569789886474609e-06],"num_classes":4},{"index":1,"label":"AMBIGUOUS","probs":[0.00013005733489990234,0.004421234130859375,0.005367279052734375,0.990234375],"num_classes":4}],"usage":{"prompt_tokens":17,"total_tokens":17,"completion_tokens":0,"prompt_tokens_details":null}}

[github-actions]

👋 Hi! Thank you for contributing to the vLLM project.

💬 Join our developer Slack at https://slack.vllm.ai to discuss your PR in #pr-reviews, coordinate on features in #feat- channels, or join special interest groups in #sig- channels.

Just a reminder: PRs would not trigger full CI run by default. Instead, it would only run fastcheck CI which starts running only a small and essential subset of CI tests to quickly catch errors. You can run other CI tests on top of those by going to your fastcheck build on Buildkite UI (linked in the PR checks section) and unblock them. If you do not have permission to unblock, ping simon-mo or khluu to add you in our Buildkite org.

Once the PR is approved and ready to go, your PR reviewer(s) can run CI to test the changes comprehensively before merging.

To run CI, PR reviewers can either: Add ready label to the PR or enable auto-merge.

🚀

[gemini-code-assist]

Summary of Changes

Hello @nie3e, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

I've summarized this pull request which adds support for the GPT2ForSequenceClassification model. This enables users to run GPT-2 models fine-tuned for sequence classification tasks directly within vLLM, leveraging its optimized inference capabilities.

Highlights

• New Model Class: I see the introduction of GPT2ForSequenceClassification, a new class inheriting from nn.Module that wraps the base GPT2Model and adds classification-specific layers.
• Classification Logic: The new class includes a pooler method that implements last-token pooling, applies a linear scoring layer, and a Softmax activation to produce classification scores.
• Model Registration: The GPT2ForSequenceClassification model is registered in the model registry, making it discoverable and loadable by vLLM.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point in your pull request via creating an issue comment (i.e. comment on the pull request page) using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands.

Feature	Command	Description
Code Review	/gemini review	Performs a code review for the current pull request in its current state.
Pull Request Summary	/gemini summary	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in issue comments and review comments.
Help	/gemini help	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist is currently in preview and may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments to provide feedback.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Footnotes

1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configureGemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

[gemini-code-assist]

Code Review

This pull request successfully adds support for GPT2ForSequenceClassification. The implementation is clear and aligns well with vLLM's architecture. The pooling mechanism correctly utilizes the last token's representation for classification. One minor improvement is suggested for the nn.Softmax layer initialization to ensure explicitness and future compatibility.

[nie3e]

@Isotr0py You are right! Thanks to your help the code looks much simpler!

[Isotr0py]

LGTM! Thanks for your patience!

[nie3e]

@Isotr0py one thing before I fix tests. I am not sure if this model is correctly put into _CROSS_ENCODER_MODELS. What do you think?

Edit: _TEXT_GENERATION_MODELS might be good place.

[Isotr0py]

You can put it in _EMBEDDING_MODELS.

[mergify]

This pull request has merge conflicts that must be resolved before it can be
merged. Please rebase the PR, @nie3e.

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

[mergify]

This pull request has merge conflicts that must be resolved before it can be
merged. Please rebase the PR, @nie3e.

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

[nie3e]

@Isotr0py Can we somehow rerun v1-test?

Timed Out
Waited 1m 30s
·
Ran in 3h 0m

[Isotr0py]

#19872 should fix the CI deadlock, you can merge from main to get the fix.

[DarkLight1337]

Can you open another PR to update the Supported Models page?