Model Conversion

1. YOLO26 (Benchmark Testing)

The Ultralytics team conducted benchmark testing of YOLO26 on ten different model formats, measuring speed and accuracy. These formats include: PyTorch, TorchScript, ONNX, OpenVINO, TF SavedModel, TF GraphDef, TF Lite, MNN, NCNN, ExecuTorch. The benchmark tests were run on Raspberry Pi 5 with FP32 precision, default input image size of 640.

Comparison Chart

2. Model Conversion

Using YOLO26's export mode for the first time will automatically install some dependencies, please wait for automatic completion!

2.1. CLI: pt → onnx, pt → ncnn

Convert PyTorch format models to onnx and ncnn

xxxxxxxxxx
cd /home/sunrise/ultralytics26

​x
yolo export model=yolo26n.pt format=onnx
# yolo export model=yolo26n-seg.pt format=onnx
# yolo export model=yolo26n-pose.pt format=onnx
# yolo export model=yolo26n-cls.pt format=onnx
# yolo export model=yolo26n-obb.pt format=onnx
​
yolo export model=yolo26n.pt format=ncnn
# yolo export model=yolo26n-seg.pt format=ncnn
# yolo export model=yolo26n-pose.pt format=ncnn
# yolo export model=yolo26n-cls.pt format=ncnn
# yolo export model=yolo26n-obb.pt format=ncnn

2.2. Python: pt → onnx → ncnn

Convert PyTorch format models to TensorRT: conversion process will automatically generate ONNX model

xxxxxxxxxx
cd /home/sunrise/ultralytics26/yahboom_demo

xxxxxxxxxx
python3 model_pt_onnx_ncnn.py

xxxxxxxxxx
from ultralytics import YOLO
​
# Load a YOLO PyTorch model
model = YOLO("/home/sunrise/ultralytics26/yolo26n.pt")
# model = YOLO("/home/sunrise/ultralytics26/yolo26n-seg.pt")
# model = YOLO("/home/sunrise/ultralytics26/yolo26n-pose.pt")
# model = YOLO("/home/sunrise/ultralytics26/yolo26n-cls.pt")
# model = YOLO("/home/sunrise/ultralytics26/yolo26n-obb.pt")
​
# Export the model to ONNX format
model.export(format="onnx")  # This will create 'yolo11n.onnx' in the same directory
​
# Export the model to NCNN format
model.export(format="ncnn")  # creates 'yolo11n_ncnn_model'
​

Note: Converted model files are located at the converted model file location

4. Model Prediction

CLI Usage

CLI currently only supports calling USB cameras. CSI camera users can directly modify the previous Python code for ONNX and NCNN model calls!

xxxxxxxxxx
cd /home/sunrise/ultralytics26

xxxxxxxxxx
yolo predict model=yolo26n.onnx source=0 save=False show

xxxxxxxxxx
yolo predict model=yolo26n_ncnn_model source=0 save=False show

References

Quick Start Guide: Raspberry Pi with Ultralytics YOLO26