Skip to content

AI platform ready to hack and to try new algorithms.

Notifications You must be signed in to change notification settings

ajakovac/NeuronLearning

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

24 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

NeuronLearning

Verbose AI platform ready to hack and try new algorithms.

Main goal: write a tunable network manager where implementation of new algorithms is easy through writing own lambda functions. Project is writen in C++

Installation:

  • clone the github content to some directory
  • install the datasets; currently three datasets are used in some ways in the programs: mnist, cifar-10-binary and stl-10
  • create a directory at DATASET_DIR, create mnist, cifar-10-batches-bin and stl10_binary subdirectories. Extract the downloaded file content to these libraries
  • in case of mnist, please rename:
    • train-images-idx3-ubyte.gz -> train_images
    • train-labels-idx1-ubyte.gz -> train_labels
    • t10k-images-idx3-ubyte.gz -> test_images
    • t10k-labels-idx1-ubyte.gz -> test_labels
  • alternatively, edit the include/cmake_variables.hpp.in file, and provide explicitly the names of the files
  • run cmake with a switch: cmake -DDATASET_DIR="" .
  • run make or build the executables with cmake --build -DDATASET_DIR="" .

About

AI platform ready to hack and to try new algorithms.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published