Polygraphy

This page includes the Python API reference documentation for Polygraphy. Polygraphy is a toolkit designed to assist in running and debugging deep learning models in various frameworks.

For installation instructions, examples, and information about the CLI tools, see the GitHub repository instead.

For a conceptual overview of the Python API, see this page.

Warning

Any APIs not documented here should be considered internal only and do not adhere to the deprecation policy for public APIs. Thus, they may be modified or removed at any time without warning. Avoid using undocumented APIs!

API Reference: Main

• Backends
  • Base Interface
    • Loaders
    • Runners
  • Common
    • Loaders
  • ONNX
    • Loaders
  • ONNX-Runtime
    • Loaders
    • Runners
  • Plugin Reference
    • Runners
  • TensorFlow 1.X
    • Loaders
    • Runners
  • TensorRT
    • Algorithm Selector
    • Calibrator
    • Config
    • Loaders
    • Optimization Profile
    • Runners
    • Utilities
• Comparing Results
  • Comparator
  • Data Structures
  • Comparison Functions
  • Postprocessing Functions
  • Data Loader

API Reference: Miscellaneous

• Common
  • Data Structures
• Global Configuration
• Constants
• CUDA Wrapper
• Data Types
• Exceptions
• Function Helpers
• JSON Utilities
• Logger
• Module Helpers
  • Import Helpers

API Reference: Development

• Command-line Tool APIs
  • Command-line Argument Groups
    • Base Interface
    • Backends
      • ONNX
      • ONNX-Runtime
      • Plugin Reference
      • TensorFlow
      • TensorRT
    • Comparator
      • Comparator
      • Comparison Functions
      • Data Loader
      • Postprocessing Functions
    • Logger
    • Model
  • Script Interface