Skip to main content

vllm_plan

Function vllm_plan 

Source
pub fn vllm_plan(
    cfg: &Qwen38Cfg,
    facts: &RduChipFacts,
    tp: usize,
) -> Result<VllmPlan, String>
Expand description

The plan the fit implies for the vendor’s vLLM backend at the model’s full context.