Julia implementation of Liquid Neural Networks - anyone wanna help make one?

Specific Domains Machine Learning
1

Following a thread on machine-learning on Julia slack, i’m interested in putting together a Julia implementation of “Liquid Neural Networks”. These are time-adaptive neural networks, inspired by neuroscience models, where each neuron’s activity is a differential equation. The synapses are also plastic and continue to adapt after training. I think this is very interesting, and Julia has a great ecosystem for automatic differentiation with differential equations.

Anyone interested in teaming together to cook up a Julia implementation?

8 Likes
2

Count me in

3

I’m also interested.

4

Uhh! I’ll try if I can. I know Tensorflow and PyTorch but haven’t learned Flux yet.

5

Hello @afishy !

Did you make any progress on that? Anyone knows any implementation working? (examples/tutorials would also be great)

Thanks a lot!

6

Hello - Has any progress been made here or should I start from scratch?

7

I’ve uploaded an implementation as a public Pluto notebook to JuliaHub. Search for LTC and it should pop up. It’s tested on the sequential MNIST benchmark mentioned in the Hasani et al. (2021) paper.

2 Likes