README.md

darkon-examples

Examples and use cases for darkon

Install python package

pip install darkon

darkon.Influence

mnist, cnn

Influence score calculation example It first trains CNN model for MNIST data, and calculate influence score to find most helpful and harmful training samples. Then, retrain the model after removing 100 most helpful training samples and 100 most harmful training samples, respectively.

• code

cifar10, resnet

• dataset: cifar10
• network: resnet110
• Upweight influence function
  • example code
  • html view
• Mislabel detection with all of layers
  • example code
  • html view
• Mislabel detection with one top layer
  • example code
  • html view

darkon.Gradcam

ImageNet, resnet

• model: used pre-trained model in TF slim
• network: resnet v1 50 in TF slim
• Gradcam & Guided Gradcam
  • example code
  • html view

Sentence polarity dataset, Text sentiment classification by CNN

• model: used pre-trained model
• network: cnn text classification by dennybritz
• Gradcam for text sentiment classification
  • example code
  • html view