GitXplorerGitXplorer
j

mlStax

public
1 stars
1 forks
1 issues

Commits

List of commits on branch master.
Unverified
066f8fd69247302898de2fabcbccedbd048619be

Seeing convergence with C++ mlstax library

jjhallard committed 9 years ago
Unverified
3f37fb48d9a8a4ad9b11994d9d0941664b7043d7

More progress on forward prop.

jjhallard committed 9 years ago
Unverified
e33fc1f685428dc61ebe5abbaca387857d98eda6

Added tests for inits and activations, all is working

jjhallard committed 9 years ago
Unverified
7b76b9535c5b99594aec28b6b47639b44d5e4284

Filling out activation and initializer hierarchies.

jjhallard committed 9 years ago
Unverified
9fe2ad25a70c05e84049c0460a063a4c26b1b528

flushed out more bugs

jjhallard committed 9 years ago
Unverified
48ce444312b4c089c102b8875af3aa6508c479af

Adding Dense::feed method and helpers

jjhallard committed 9 years ago

README

The README file for this repository.

mlStax

Stackable deep learning models in Python, C++, and Haskell.

A senior project for CS112 Comparative Programming Languages at UCSC, Winter 2016. License - MIT

This project provides 3 similarly-structured deep-learning libraries, in the Python, C++, and Haskell programming languages. The user creates ML models by stacking customizable layers on top of one another. The user can then train the custom model over arbitrary data-sets, and if satisfied, use that model for predictive purposes.

I do plan on adding in CUDA bindings for each library, but that's definitely for v.2.