What do we expect from a candidate?
- Building deep learning models
- Meticulously designing and executing experiments to validate the thesis
- Analyzing large amounts of ambiguous data
We are looking for a person who is creative, inquisitive and questions the results of their experiments. Preferably someone who will not be afraid to ask difficult questions.
- PyTorch, PyTorch Lightning
- hydra (config system) – we do not require knowledge of hydra specifically but familiarity with running experiments using the yaml config system
Day to day
Meetings at 9:00 am where we discuss the research currently underway. Open discussion where each team member shows the progression of their research in the form of reports on wandb and we discuss conclusions together. Most of the work consists of:
- Drafting the research process – from asking the question we want to solve, to planning each individual experiment
- Programming custom solutions to let go of specific experiments
- Designing the architecture of neural networks
- Analyzing the experiments