Skip to content

Latest commit

 

History

History
12 lines (9 loc) · 531 Bytes

README.md

File metadata and controls

12 lines (9 loc) · 531 Bytes

Classify

It is a simple handwriting digits recognition based on CNN.

The CNN folder contains:

  1. CNN_example.py: use to test single image
  2. CNN_test.py: use to measure the quality on test dataset
  3. CNN3_train.py: use to train the model
  4. my_net: the trained model

The same as DNN and resnet. By the way, the "tf_logs_test" and "tf_logs_train" are used for visualization that are generated from the project.

It implement on tensorflow 1.8.0, and the tensorboard is also used to visualize the taining process.