Prep tasks for contributing to Tesla once hired.
At the very least:
- Python: coding, debugging, profiling, version control.
- PyTorch (or TensorFlow)
- Large-scale distributed clusters: CPU/GPU interactions/transfers, latency/throughput bottlenecks during training, CUDA, pipelining/multiprocessing.
- DL fundamentals, SOTA models as applied to vision (segmentation, detection) (or speech, NLP).
- SOTA paper implementation/replication
- Data science tools: Python scripting, numy, scipy, matplotlib, scikit-learn, jupyter, bash, Linux.
- (Neural net deployment on embedded systems)
Based on this job announcement.