PopEF: User Guide

• 1. Introduction
• 2. PopEF file format
  • 2.1. Blob header
  • 2.2. Poplar executable
  • 2.3. Opaque
  • 2.4. Metadata
    • 2.4.1. Target runtime
    • 2.4.2. Program flow
    • 2.4.3. Anchors
  • 2.5. Tensor and feed data
  • 2.6. PopEF file analysis
• 3. PopEF library usage
  • 3.1. Writing a PopEF stream
  • 3.2. Multi-feed writers
  • 3.3. Reading PopEF streams
• 4. Model handling
  • 4.1. Exporting models
    • 4.1.1. General approach
    • 4.1.2. How to export
    • 4.1.3. Poplar Executable
  • 4.2. Importing models
    • 4.2.1. Model management
• 5. PopEF C++ API reference
  • 5.1. Metadata
  • 5.2. Classes for writing PopEF streams
  • 5.3. Classes for reading PopEF streams
  • 5.4. Model representation classes
  • 5.5. Error classes
• 6. PopEF Python API reference
  • 6.1. PopEF Metadata
  • 6.2. Classes for writing PopEF streams
  • 6.3. Classes for reading PopEF streams
  • 6.4. Model representation classes
• 7. Environment variables
  • 7.1. POPEF_LOG_DEST
  • 7.2. POPEF_LOG_LEVEL
• 8. Trademarks & copyright