# Coreml Profile

> Profile Kokoro and related Core ML packages to establish actual CPU, GPU, and Neural Engine placement, latency, and silent fallback behavior. Use for device compute plans, scheduling claims, package-shape changes, warmed timing, or hardware comparisons.

- Skill: `mattmireles/coreml-profile` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add mattmireles/coreml-profile`
- Raw SKILL.md: https://api.skillmd.com/api/skills/mattmireles/coreml-profile/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: mattmireles (https://skillmd.com/u/mattmireles)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/mattmireles/coreml-profile

---


# Profile Kokoro Core ML

Bind every measurement to the package hash, model/configuration, hardware, OS,
Core ML compute-unit request, input shape/duration, warmup policy, and timing
boundary. Inspect a device compute plan or physical trace when available; never
infer Neural Engine placement from a requested compute unit or speed alone.

Start with `scripts/dump_device_compute_plan.py`,
`scripts/inspect_coreml_compute_plan.m`, and
`README/Guides/apple-silicon/CoreML-Compute-Unit-Scheduling-guide.md`.
Keep synthesis, phonemization, I/O, and individual model-stage timing separate.