Training from scratch

Edit config.yml as below.

finetune:
  models:
    - scratch-alexnet

You can also run fine-tuning and training from scratch together.

finetune:
  models:
    - imagenet1k-inception-v3
    - scratch-inception-v3

Available models traininig from scratch

from scratch model name
scratch-alexnet
scratch-googlenet
scratch-inception-bn
scratch-inception-resnet-v2
scratch-inception-v3
scratch-inception-v4
scratch-lenet
scratch-mlp
scratch-mobilenet
scratch-resnet-N
scratch-resnext-N
scratch-vgg-N

Specify the number of layers for N in scratch-resnet, scratch-resnext and scratch-vgg.

For scratch-resnet and scrach-resnext, N can be set to 18, 34, 50, 101, 152, 200 and 269, and for scratch-vgg, N can be set to 11, 13, 16 and 19.