Implementation of efficient, scalable deep learning models
Scalable Machine Learning
Description
BACKGROUND
Event-based neural network models are biologically inspired [1] and have the ability to scale well [2] while being extremely energy efficient on neuromorphic hardware [5].
YOUR TASK
Create a high-performance implementation of event-based models based on [6]. This can be using one of the following
- Rust or Elixir
- As an extension of GeNN
- Using MPI in C++
- For the Graphcore AI chip.
Required Skills
- Good Knowledge of Rust/Elixir/C++/CUDA depending on the choice of implementation.
- Knowledge of concurrent programming concepts.
- Basic Knowledge in deep learning.
References
[1] Pfeiffer, Michael, and Thomas Pfeil. "Deep learning with spiking neurons: opportunities and challenges." Frontiers in neuroscience (2018): 774. (link)
[2] Jordan, J., Ippen, T., Helias, M., Kitayama, I., Sato, M., Igarashi, J., Diesmann, M. and Kunkel, S., 2018. Extremely scalable spiking neuronal network simulation code: from laptops to exascale computers. Frontiers in neuroinformatics, p.2. (link)
[3] NEST Simulator
[4] GeNN, a GPU-enhanced Neuronal Network simulation
[5] Christensen, Dennis Valbjørn, et al. "2022 roadmap on neuromorphic computing and engineering." Neuromorphic Computing and Engineering (2022).
[6] Subramoney, A., Nazeer, K.K., Schöne, M., Mayr, C., Kappel, D., 2022. EGRU: Event-based GRU for activity-sparse inference and learning. https://doi.org/10.48550/arXiv.2206.06178