SqueezeNet1_1

Quantized SqueezeNet1_1 model that could be supported by AMD Ryzen AI.

Model description

SqueezeNet was first introduced in the paper SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size. This model is SqueezeNet v1.1 which requires 2.4x less computation than SqueezeNet v1.0 without diminshing accuracy.

The model implementation is from torchvision.

How to use

Installation

Follow Ryzen AI Installation to prepare the environment for Ryzen AI. Run the following script to install pre-requisites for this model.

pip install -r requirements.txt 

Data Preparation

Follow PyTorch Example to prepare dataset.

Model Evaluation

python eval_onnx.py --onnx_model SqueezeNet_int8.onnx --ipu --provider_config Path\To\vaip_config.json --data_dir /Path/To/Your/Dataset

Performance

Metric	Accuracy on IPU
Top1/Top5	57.70% / 80.27%

 @article{SqueezeNet,
    Author = {Forrest N. Iandola and Song Han and Matthew W. Moskewicz and Khalid Ashraf and William J. Dally and Kurt Keutzer},
    Title = {SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and $<$0.5MB model size},
    Journal = {arXiv:1602.07360},
    Year = {2016}
}