/ImageClassification_Benchmark

Image classification codes based on Pytorch1.4 and Tensorflow2.0

Primary LanguageJupyter NotebookApache License 2.0Apache-2.0

ImageClassification_Benchmark

Stars Forks

Image classification workflow based on Pytorch 1.4 and Tensorflow 2.

Dataset

MNIST, CIFAR10, CIFAR100 and ImageNet are supported.

Features

Tensorboard, distributed training, mixed precision training, quantization and more feature are coming soon

Log

  • 24 Mar Update:
    • Update training/evaluation code

Todo list

  • Pytorch
    • Dataset ,dataloader, preprocessing and visualization
    • Classical & state-of-the-art models(AlexNet, VGGNet, Inceptions, ResNet, ResNeXt, DenseNet, SeNet, MobileNet, e.t.c)
    • Training and Evaluation codes
    • Tensorboard support
    • Binary quantization models
    • Distributed training
    • Arbitrary bits quantization
  • Tensorflow
    • Dataset ,dataloader, preprocessing and visualization
    • Classical & state-of-the-art models(AlexNet, VGGNet, Inceptions, ResNet, ResNeXt, DenseNet, SeNet, MobileNet, e.t.c)
    • Training and Evaluation codes
    • Tensorboard support
    • Binary quantization models
    • Distributed training
    • Arbitrary bits quantization
  • Summary
    • Mannual
    • Result benchmarks

Reference

  1. https://github.com/eladhoffer/convNet.pytorch
  2. https://github.com/wonnado/binary-nets
  3. https://github.com/666DZY666/model-compression