Skip to content
master
Switch branches/tags
Go to file
Code

Latest commit

 

Git stats

Files

Permalink
Failed to load latest commit information.
Type
Name
Latest commit message
Commit time
 
 
src
 
 
 
 
 
 

SmallTrain

The machine learning library enabling "small train", which requires lower machine power and fewer training data.

Table of Contents

Overview

SmallTrain is a machine learning library running on TensorFlow (and Keras and PyTorch in the future version). It enables small train, which requires lower machine power and fewer training data because you can develop your machine learning(ML) models from our pre-trained models.

Features

  • Easy to develop for POC to production
  • Almost no programming for building your pre-trained model.
  • Available as both TensorFlow and Pytorch wrapper.
  • Always adapting to algorithms which evolve.
  • Build using state-of-the-art algorithms from Scientific and Mathematical papers
  • Accuracy is always going to be better even with minimal data and training time
  • Licensed under MIT Open Source

Docs & Community

Installation

Standalone Docker Application

See this guide for detailed instruction to build SmallTrain as a standalone docker container from source.

Python Library

This is a SmallTrain module available through the pypi registry.

Installation is done using the pip install command:

$ pip install smalltrain

Follow our installing guide for more information.

Tutorials

You can run tutorial codes on Jupyter Lab Notebook.

Launch Jupyter Lab Notebook

# Enable password
jupyter notebook password
# Run Jupyter Lab Notebook
cd /var/smalltrain/tutorials
nohup jupyter lab &

For example, a tutorial notebook for image detection is available on http://YOURHOST:JUPYTER_NOTEBOOK_PORT/lab/tree/image_recognition/notebooks/cifar10.ipynb (Default JUPYTER_NOTEBOOK_PORT is 8888).

See Tutorials for more tutorials.

Contributing

Contributing Guide

List of all contributors

Licence

MIT Licence

About

No description, website, or topics provided.

Resources

License

Releases

No releases published

Packages

No packages published