metadata

license: apache-2.0
datasets:
  - imagenet-1k
metrics:
  - accuracy
tags:
  - RyzenAI
  - vision
  - classification
  - pytorch

Inception_v4

Quantized Inception_v4 model that could be supported by AMD Ryzen AI.

Model description

Inception_v4 was first introduced in the paper Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning.

The model implementaion is from TensorFlow-Slim.

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 imagenet-1k to download dataset.

Download ImageNet validation synset labels file.

Create validation image list:

python create_image_list.py imagenet_2012_validation_synset_labels.txt

Model Evaluation

python eval_onnx.py --onnx_model inceptionv4_int8.onnx --ipu --provider_config Path\To\vaip_config.json --val_data_dir /Path/To/Your/Validation/Data --val_image_list val.txt

Performance

Metric	Accuracy on IPU
Top1/Top5	79.92% / 95.02%

@article{Szegedy2016Inceptionv4IA,
  title={Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning},
  author={Christian Szegedy and Sergey Ioffe and Vincent Vanhoucke and Alexander A. Alemi},
  journal={arXiv:1602.07261},
  year={2016},
}