$ show onnx/models
This repository is a deprecated ONNX Model Zoo containing thousands of pre-trained models (mostly computer vision, some NLP and generative AI) exported to the ONNX format from PyTorch (timm, torchvision) and other frameworks. It also contains turnkeyml benchmark artifacts (turnkey_stats.yaml) that record export success, model metadata, ONNX operator counts, and system environment for each model across opsets 16–18. The primary purpose was to provide a centralized collection of ready-to-use ONNX models for inference and to validate cross-version ONNX export compatibility. However, the repository is explicitly deprecated (LFS disabled July 2025) and users are redirected to Hugging Face. The codebase consists almost entirely of binary model files and YAML metadata; the only reusable software components are a few preprocessing scripts (e.g., imagenet_preprocess.py, face_preprocess.py) in the validated/ directories. There is no application logic, no API, no training pipelines—this is purely a model distribution and benchmark data repository. Reproducibility is low due to hardcoded paths, frozen env snapshots that are out of date, and reliance on external model sources whose weights may have changed.
The LLM identified 2 files worth highlighting for reuse:
A self-contained Python script that resizes and normalizes images for ImageNet-style classification models. It applies standard preprocessing (resize to 256, center crop to 224, normalize with mean and std) and returns a numpy array ready for ONNX Runtime inference. This script can be extracted into any vision inference project with minimal changes to the input/output paths.
A standalone utility for face alignment and preprocessing used by face recognition models (e.g., ArcFace). It implements similarity transform based on detected landmarks and resizes/crops to a fixed size (112x112). The core alignment logic is model-agnostic and can be reused with any face recognition or verification pipeline that requires aligned face crops.