/TensorRT-solo-python

TensorRT for SOLO(use python)

Primary LanguagePythonOtherNOASSERTION

TensorRT SOLO (Python)

996.icu

TensorRT for SOLO(use python)

Enviroments

TensorRT >=7.1
Ubuntu 18.04

A quick demo

1. Convert solo model form pytorch to onnx

python3 get_onnx.py --config ${SOLO_path}/configs/solov2_r101_fpn_8gpu_3x.py --checkpoint ${SOLO_path}/work_dirs/SOLOv2_R101_3x.pth --outputname solov2_r101.onnx

2.Genrate tensorRT engine and inference

python3 inference.py --onnx_path solov2_r101.onnx --engine_path solov2_101.engine --mode fp16 --image_path ${your_picture_path} --save --show

Inference performance(only inference time in GPU)

GPU Model Mode Inference time
V100 solov2 r101 fp16 35ms
Xavier solov2 r101 fp16 150ms