chore (1.x): support vllm 0.18.0 without backwards compat

[tjohnson31415]

Description

We want to make another release on the 1.x line and need to use the latest vLLM. We are allowing this change to drop support for older vLLM versions. vLLM supported range will now have 0.17.0 as a lower bound.

Related Issues

Test Plan

Must pass all standard CI tests.

Checklist

• I have read the contributing guidelines
• My code follows the project's code style (run bash format.sh)
• I have added tests for my changes (if applicable)
• I have updated the documentation (if applicable)
• My commits include a Signed-off-by: line (DCO compliance)

[github-actions]

👋 Hi! Thank you for contributing to vLLM support on Spyre.
Just a reminder: Make sure that your code passes all the linting checks, otherwise your PR won't be able to be merged. To do so, run ./format.sh.
Now you are good to go 🚀.

We also recommend installing prek and configuring it to check your code before every local commit.

[maxdebayser]

LGTM

[tjohnson31415]

Had a job fail with an client timeout / broken connection a couple times, but now it seems to have passed. I also reproduced it locally, but the was unable to reproduce it after a few more tries. So we may have a little bit of flakiness 🤔

[tjohnson31415]

bot:test
MARKERS="spyre and not quantized"

[tjohnson31415]

bot:test
MARKERS="spyre and prefix_caching and not quantized"

[tjohnson31415]

bot:test
MARKERS="spyre and cb and not quantized"

[tjohnson31415]

bot:test
MARKERS="spyre and cb and not quantized"

[tjohnson31415]

Remaining spyre-ci failures have been confirmed to be unrelated to this PR's changes.