wang-xinyu/tensorrtx

What It Solves

Engineers deploying deep learning models on NVIDIA GPUs need low-latency inference; TensorRT requires manual network definition. tensorrtx provides reference implementations for popular architectures using TensorRT's C++ API.

Primary use cases

• Deploying YOLO11, DETR, ResNet, MobileNetV2/V3, Swin-Transformer on TensorRT
• Porting PyTorch models to TensorRT via reference implementations

Secondary use cases

• Benchmarking TensorRT vs ONNX Runtime/OpenVINO
• Developing custom TensorRT plugins

When to Use

Consider when

• Target hardware is NVIDIA GPU
• Maximum inference throughput required
• Model architecture is in supported list (YOLO11, DETR, ResNet, MobileNet, Swin, etc.)

Verify before adopting

• ? TensorRT version compatibility with your CUDA/cuDNN
• ? Numerical accuracy after FP16/INT8 optimization
• ? Maintenance status of specific model implementation

Start Here

Adoption Checklist

• Needs verification TensorRT version compatibility with your CUDA/cuDNN
• Needs verification Numerical accuracy after FP16/INT8 optimization
• Needs verification Maintenance status of specific model implementation

Each check stays "needs verification" until an official source confirms it; unconfirmed items are never marked verified.

Known Limitations & Unknowns

Known limitations

• NVIDIA GPU only
• Manual network definition (no automatic ONNX parser)
• Limited to architectures implemented in repo
• INT8 calibration not included for all models

Alternatives & Related Tools

Related tools

• tensorrt — Built on
• detr — Implements

How is it used?

Start from the recorded entry points below, then validate against the technical checklist.

• Official site https://github.com/wang-xinyu/tensorrtx
• Repository https://github.com/wang-xinyu/tensorrtx
• Documentation https://github.com/wang-xinyu/tensorrtx

Technical checklist

• OK License identified Recorded: MIT
• OK Maintenance signal Active
• OK Verification status Official confirmed
• OK Source evidence attached 1 source record(s)
• NEEDS REVIEW Latest version recorded Not recorded

Official Links

Metadata & Governance

Dataset facts

Facts above come from the official dataset card only; unconfirmed fields stay unknown.

Related Resources & Dependencies

• tensorrt — built on (confirmed)
• detr — implements (confirmed)

Recent Activity

Related Knowledge

Guides

• Deploying AI Inference on UAV Edge Compute — A repeatable workflow for choosing and validating an inference runtime for an onboard UAV computer.

Collections

• Edge AI Perception Starter — A starting stack for prototyping aerial perception models before edge deployment work.