fx

Step-by-Step

This document describes the step-by-step instructions for reproducing PyTorch se_resnext tuning and benchmarking results with Intel® Neural Compressor.

Prerequisite

1. Environment

Python 3.6 or higher version is recommended. The dependent packages are all in requirements, please install as following.

cd examples/pytorch/image_recognition/se_resnext/quantization/ptq/fx
pip install -r requirements.txt

2. Install from Repo

git clone https://github.com/Cadene/pretrained-models.pytorch.git
cd pretrained-models.pytorch
git checkout 8aae3d8f1135b6b13fed79c1d431e3449fdbf6e0
python setup.py install
cd ..

Note

Please don't install public pretrainedmodels package.

3. Prepare Dataset

Download ImageNet Raw image to dir: /path/to/imagenet. The dir should include below folder:

ls /path/to/imagenet
train  val

Run

1. Quantization

python run_eval.py \
          --data /path/to/imagenet \
          -a se_resnext50_32x4d \
          -b 128 \
          -j 1 \
          -t

2. Benchmark

# int8
sh run_benchmark.sh --int8=true --config=saved_results --mode=performance --input_model=se_resnext50_32x4d  --dataset_location=/path/to/imagenet
# fp32
sh run_benchmark.sh --mode=performance --input_model=se_resnext50_32x4d  --dataset_location=/path/to/imagenet

Original SE_ResNext README

Please refer SE_ResNext README.