vllm_omni.diffusion.sched.request_scheduler ¶

RequestScheduler ¶

Bases: _BaseScheduler

Diffusion scheduler with vLLM-style waiting/running queues.

add_request(request: OmniDiffusionRequest) -> str

schedule() -> DiffusionSchedulerOutput

update_from_output(
    sched_output: DiffusionSchedulerOutput,
    output: RunnerOutput,
) -> set[str]