Inference Tune Sweep

Search a vLLM serving model's config space for its best inference performance: a `perftunereport campaign_run` matrix sweep over concurrency x max_num_batched_tokens x kv-cache-dtype x cudagraph_mode, plus a same-node controlled A/B (`run-controlled-ab.sh`) for any decode-latency claim, gated by `perf-baseline-diff`. Picks a champion config that beats the baseline on the run's focus metric (throughput tok/s OR median TPOT/ITL), honoring warm/cold + eager/cudagraph methodology rules. This is the tuning phase of `inference-model-optimize`, usable standalone. Not a Bayesian optimizer -- it sweeps an operator-named grid (an --optimizer hook is stubbed). Triggers on "tune the vllm config", "sweep max_num_batched_tokens", "find the best serving config", "vllmArgs sweep", "cudagraph_mode A/B", "kv-cache-dtype sweep", "tune concurrency", "config search for <model>", or any combination of "tune / sweep / optimize / search / A-B" with "vllm / config / vllmArgs / concurrency / batched-tokens / cudagraph".

mkdir -p ~/.claude/skills/inference-tune-sweep
curl -L https://claudskills.com/skills/inference-tune-sweep/SKILL.md \
  -o ~/.claude/skills/inference-tune-sweep/SKILL.md

@misc{inference-tune-sweep-2026,
  author    = {cfregly},
  title     = {Inference Tune Sweep [Claude Code skill]},
  year      = {2026},
  publisher = {ClaudSkills},
  url       = {https://claudskills.com/skills/inference-tune-sweep/}
}