Phase 6 · Production

vLLM, Deeply

Not 'can use it' — 'has read it'.

0/4 · 600 XP

You’re previewing the curriculum. Sign in (top right) to track progress, take knowledge checks, and unlock the verifiers.

Tasks

Read “Inside vLLM: Anatomy of a High-Throughput Inference System”
The closest thing to a textbook chapter on engine internals. Study the V1 architecture — V0 posts describe a dead engine.
read+100 XPresource ↗
Trace one request through the vLLM source
Arrival → scheduler → block manager → model runner → sampler → stream. Write down every file and class it touches.
read+150 XP
Deploy vLLM on the fleet and tune it
max-num-batched-tokens, prefix caching, chunked prefill, quantized weights + FP8 KV. One knob at a time, keep a results table.
build+150 XPresource ↗
VERIFIED: your production vLLM endpoint passes the conformance probe
Expose it (with an API key) and let InferQuest probe it live — same suite your toy engine passed, now against production infrastructure.
build+200 XPauto-verified