New Features: NVIDIA GPU Inference support via TensorRT
🎉🎉TensorRT: Boost inference performance with NVIDIA GPU via TensorRT.
Run bash ./build.sh tensorrt
to build lite.ai.toolkit with TensorRT support, and then test yolov5 with the codes below. NOTE: lite.ai.toolkit need TensorRT 10.x (or later) and CUDA 12.x (or later). Please check build.sh, tensorrt-linux-x86_64-install.zh.md, test_lite_yolov5.cpp and NVIDIA/TensorRT for more details.
// trtexec --onnx=yolov5s.onnx --saveEngine=yolov5s.engine
auto *yolov5 = new lite::trt::cv::detection::YOLOV5(engine_path);
Class | Class | Class | Class | Class | System | Engine |
---|---|---|---|---|---|---|
✅YOLOv5 | ✅YOLOv6 | ✅YOLOv8 | ✅YOLOv8Face | ✅YOLOv5Face | Linux | TensorRT |
✅YOLOX | ✅YOLOv5BlazeFace | ✅StableDiffusion | / | / | Linux | TensorRT |
What's Changed
- Add Yoloface-8n by @wangzijian1010 in #412
- [README] Update authors by @DefTruth in #413
- 增加Linux下的TensorRT的安装文档 by @wangzijian1010 in #415
- update first tensortt inference example by @wangzijian1010 in #416
- Update TensorRT BasicHandler and add yolov5 cpp file by @wangzijian1010 in #417
- add yolox implement by @wangzijian1010 in #418
- [TensorRT] support YOLOv8 with TensorRT backend by @wangzijian1010 in #419
- [TensorRT] support YOLOv6 with TensorRT backend by @wangzijian1010 in #420
- [TensorRT] support YOLOV5Face with TensorRT backend by @wangzijian1010 in #421
- [TensorRT] fix undefine bug in yolov5face trt version by @wangzijian1010 in #422
- [onnxruntime] Clip onnx update by @wangzijian1010 in #424
- [Misc] Remove un-need comments by @wangzijian1010 in #425
- [TensorRT] Clip TensorRT code update by @wangzijian1010 in #426
- Update README.md by @DefTruth in #430
- [TensorRT] [ORT] support Stable Diffusion text2img Pipeline by @wangzijian1010 in #432
- [Bugfix] Fix some build errors for SD by @DefTruth in #434
- [Bugfix] Fix OrtCompatiableGetInputName errors by @DefTruth in #435
- [Fix Bug] remove duplicate code by @wangzijian1010 in #436
- Bump up to v0.3.0 by @DefTruth in #438
New Contributors
- @wangzijian1010 made their first contribution in #412
Full Changelog: v0.2.0...v0.3.0-rc1