rknn-toolkit2/examples/pytorch/resnet18_qat/README.md

Pytorch ResNet18 QAT

Model Source

The model used in this example come from the 'torchvision', more details in the 'export_pytorch_model' function of the script.

Script Usage

Usage:

python test.py

rknn_convert usage:

python3 -m rknn.api.rknn_convert -t rk3568 -i ./model_config.yml -o ./

Description:

• The default target platform in script is 'rk3566', please modify the 'target_platform' parameter of 'rknn.config' according to the actual platform.
• If connecting board is required, please add the 'target' parameter in 'rknn.init_runtime'.
• This is a QAT model, and the do_quantization of rknn.build needs to be set to False.

Expected Results

This example will print the TOP5 labels and corresponding scores of the test image classification results, as follows:

-----TOP 5-----
[812] score:0.999741 class:"space shuttle"
[404] score:0.000194 class:"airliner"
[657] score:0.000015 class:"missile"
[466] score:0.000008 class:"bullet train, bullet"
[744] score:0.000008 class:"projectile, missile"

• Note: Different platforms, different versions of tools and drivers may have slightly different results.