Core AI 模型库

Rebuild a bundle — the conversion recipes

Every published bundle carries the configuration that produced it in models/<model>/recipe.toml, and one command runs it:

python3 conversion/zoo_convert.py list                    # what can be rebuilt
python3 conversion/zoo_convert.py show  qwen3.5-0.8b      # the command + everything it needs
python3 conversion/zoo_convert.py run   qwen3.5-0.8b --dry-run
python3 conversion/zoo_convert.py run   qwen3.5-0.8b      # do it

show prints four kinds of prerequisite, and the run-time ones matter as much as the rest — an export that skips them still succeeds, and the bundle then misbehaves inside the app:

Ports whose exporter is self-contained need no setup at all:

uv run conversion/export_da3.py --variant small --dtype float16 --res 504

Paths never assume a machine: python3 conversion/_paths.py prints where downloads, exports and the Hugging Face cache resolve, and ZOO_WORK_ROOT / ZOO_EXPORTS / ZOO_CODE_ROOT / HF_HUB_CACHE move them.

A recipe marked unverified refuses to run without --force. It means the repository does not record which configuration produced the published bundle, and it prints the exact question it cannot answer. Running it anyway yields a bundle, not the bundle.

Do not expect a checksum match. Conversion is not byte-deterministic here: the same recipe run twice on the same machine produces bundles that differ from each other (measured: main.mlirb by 7 bytes, main.hash entirely). A rebuild is judged by the gates the export runs — and by checking the published artifact itself:

python3 conversion/zoo_verify.py mlboydaisuke/Gemma-4-12B-CoreAI     # one repo
python3 conversion/zoo_verify.py --all --json models/_VERIFY.json    # the whole catalog, minutes

That compares a bundle's tokenizer, chat template, context length and declared precision against the source model it names in its own metadata.json — no oracle, no device, no weights.

The numerical check is a different command. zoo_verify.py checks that a bundle is described correctly; it does not check that the bundle still computes the right thing. That is conversion/coreai_gate.py, which rebuilds the reference model in fp32 and compares a greedy decode token for token:

python3 conversion/coreai_gate.py <bundle-dir> Qwen/Qwen3.5-2B --revision <sha> \
    --transcript models/qwen3.5-2b/gate.json

It runs outside this working tree — point --runner / ZOO_LLM_RUNNER at your llm-runner and --python / ZOO_CONVERT_PYTHON at an interpreter that has the export overlay, and it tells you which one is missing rather than failing obscurely. It currently covers the decode architectures in --arch (the oracle is a per-architecture transcription of that model's export, so it grows one port at a time, not all at once).

--transcript is the part worth publishing: the pinned revision, the exact input_ids, both sides' generated tokens, the tie margins, and the verdict. Rebuilding the oracle is expensive; re-running the engine side against a published transcript is not — it needs the bundle, llm-runner, and the recorded input_ids, and the output must match engine.gen_text.

New ports publish one. For models ported before the flag existed, conversion/backfill_gate_transcripts.py prints what can still be gated and, with --run, gates it — preferring the published bundle out of the CoreAIKit cache over a local export, so the transcript describes the bytes apps download at the revision the catalog pins. Nothing reconstructs a transcript from a card's prose after the fact: a model that cannot be re-gated keeps a card that says what was gated, without a file claiming to prove it.

Results land in models/_INVENTORY.md; models/index.json is the same catalog machine-readable, which is where an agent should start. Each recipe entry there carries status (does the repository record what produced this bundle) and gate_transcript (is the numerical check against the original published, and where) — two different questions, kept as two fields so neither has to be inferred from prose.

Install

Claude Code

/plugin marketplace add https://github.com/john-rocky/coreai-model-zoo
/plugin install coreai-zoo-skills@coreai-model-zoo

Codex CLI

codex plugin marketplace add https://github.com/john-rocky/coreai-model-zoo

then /plugins → coreai-model-zoo → coreai-zoo-skills → Install.

Gemini CLI

gemini extensions install /path/to/coreai-model-zoo/skills

Apple's own coreai-skills covers the toolchain itself (authoring, export, compression). These two cover this catalog; install both.

Decode throughput (tok/s, greedy; output top-1 exact vs the Hugging Face reference)

Measured on the iOS 27 / macOS 27 beta, Apple's coreai-pipelined GPU engine, zero custom kernels (ANE column + †/‡ excepted). † = MoE bundle using the custom gather_qmm Metal kernel (reads only the routed experts). ‡ = dense bundle whose full/global-attention SDPA is a custom flash-decode Metal kernel — the stock MPSGraph SDPA crashes on the ≥16-head × 512 Q (a GPU scratch-heap overflow, apple/coreai-models#27), so these models are unrunnable without it. Prefill, sizes, per-model caveats, and the Mac-only big models: models/.

<p align="center"> <img width="380" alt="CoreAIChat screen recording" src="https://github.com/user-attachments/assets/999dbd95-45b5-468f-b1a8-34112ee3b74d" /> </p> <p align="center"><i>CoreAIChat (<a href="apps/">apps/</a>) — the zoo's models running on-device on iPhone.</i></p>