Model Lint

Runs static quality checks on model metadata for best practices. Each lint rule checks a specific aspect of the model (compression, quantization, naming conventions, dtype consistency) and reports findings with severity and actionable suggestions.

CLI Equivalent

apr lint model.apr

Key Concepts

• Static quality analysis of model metadata
• Severity-based lint reporting (info, warn, error)
• Best-practice enforcement for compression, naming, dtypes

Run

cargo run --example analysis_lint

Source

examples/analysis/analysis_lint/main.rs