CLI reference¶

The PyPI package is modelport-cli. It installs the modelport command.

pip install "modelport-cli[onnx,executorch,torchvision,hf,gguf]"

Extra	Adds	Needed for
`onnx`	onnx, onnxruntime, onnxscript, torch	ONNX export, quantize, verify
`executorch`	executorch, torch	ExecuTorch export and verify
`torchvision`	torchvision, torch	`torchvision:` sources
`hf`	transformers, huggingface_hub, torch	`hf:` sources, `import-gguf` from the Hub, `publish --hf`
`gguf`	gguf	Inspecting GGUF files

Commands¶

Command	What it does
`export SOURCE`	Convert a model and write a bundle with `modelport.json`, labels, and golden data.
`quantize BUNDLE`	Add `--fp16` and `--int8` ONNX variants with measured tolerances.
`verify BUNDLE`	Run every variant on the golden input and compare with PyTorch.
`import-gguf SOURCE`	Describe GGUF language models from a Hugging Face repo or a local file.
`pack BUNDLE`	Refresh sizes and hashes after manual edits; list unlisted files.
`publish BUNDLE`	Upload with `--github owner/repo --tag tag` or `--hf org/name`.
`gen-dart MANIFEST`	Write a typed Dart wrapper with named inputs and outputs.
`inspect FILE`	Show inputs, outputs, and metadata of `.onnx`, `.pte`, or `.gguf` files.
`validate MANIFEST...`	Check manifests against the spec.
`schema`	Print the manifest JSON Schema.
`doctor`	Check Python, optional packages, and disk space.

Run modelport COMMAND --help for every option.

Sources¶

Source	Example	Notes
torchvision	`torchvision:efficientnet_b0`	Classifiers, with torchvision's own preprocessing
Hugging Face	`hf:facebook/deit-tiny-patch16-224`	Image classifiers and DETR-family detectors; local folders work too
Your code	`file:my_model.py:build`	A function returning `modelport.sources.SourceModel`

file: runs the code in that file, so only use files you trust. Load checkpoints with torch.load(..., weights_only=True) or safetensors, never plain pickle from strangers.