English

Documentation Structure

This document is intended for all developers using D-Robotics LLM and provides comprehensive guidance for the entire development process.

The following is a brief overview of each chapter. You may refer to the corresponding sections as needed.

Provides an overview of the content within each chapter of this document.

Displays the list of models supported on the D-Robotics edge side and their mapping to on-device execution examples.

Describes the release package contents supported by the D-Robotics LLM toolchain.

Chapter	Description
X86 Environment Setup	Instruction on setting up the development environment on an x86 machine.
On-device Environment Setup	Instruction on setting up the development environment on the edge device.

Chapter	Description
On-device Preparation	General preparation steps required before running models on the edge device.
LLM Development Guide	Development workflow for the LLM (Large Language Model) series.
VLM Development Guide	Development workflow for the VLM (Vision Language Model) series.
VLA Development Guide	Development workflow for the VLA (Vision-Language-Action) series.
ASR Development Guide	Development workflow for the ASR (Automatic Speech Recognition) series.
Appendix	Parameter descriptions for configuration files required for on-device execution.

Chapter	Description
Overview	Illustrates the basic API call workflow for model loading and inference.
API Details Introduction	Detailed technical specifications, including basic types, structures, enums, callbacks, and core functions.

Chapter	Description
Benchmark Specifications	Defines testing conditions and statistical terminology used in benchmarks.
LLM Benchmark	Performance metrics for the LLM series.
VLM Benchmark	Performance metrics for the VLM series.
VLA Benchmark	Performance metrics for the VLA series.
ASR Benchmark	Performance metrics for the ASR series.

The authorized user agreement. Please read this carefully before using the D-Robotics LLM toolchain.

#Documentation Structure