[docs] rephrase comment about continuous batching configuration

[yannicks1]

The previous comment was slightly misleading, as VLLM_SPYRE_WARMUP_NEW_TOKENS is an upper limit on the number of generated output tokens by the Spyre compiler for static batching, while (SamplingParams.)max_tokens is a user defined parameter which has to obey the compiler constraint for both static and continuous batching.

By rephrasing the continuous batching comment to:

The number of generated output tokens is implicitly limited by max-model-len - padded_prompt_length

both the static and continuous batching section now refer to the compiler side upper limit.

[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, first install the linting requirements, then run format.sh and commit the changes. This can be done with uv directly:

uv sync --frozen --group lint --active --inexact

Or this can be done with pip:

uv pip compile --group lint > requirements-lint.txt
pip install -r requirements-lint.txt
bash format.sh

Now you are good to go 🚀

[joerunde]

lpgtm!