You can find my article describing the levels in more detail at Towards Data Science. This list is also on GitHub.

Entry level

Data handling

• [ ] Small datasets
• [ ] Simple preprocessing
• [ ] Image data
• [ ] Text data
• [ ] Audio and time-series data

Networks

• [ ] Classic machine learning
• [ ] Basic Neural Networks
• [ ] Convolutional Neural Networks
• [ ] Recurrent Neural Networks

General

• [ ] Data analysis
• [ ] Saving and loading models
• [ ] Working with metadata files
• [ ] Callbacks

Intermediate level

Data handling

• [ ] Generators
• [ ] Augmentations
• [ ] Large datasets
• [ ] Custom pipelines

Custom projects

• [ ] Custom audio / time-series project
• [ ] Custom image project
• [ ] Custom text project

Training

• [ ] Transfer learning
• [ ] Fine-tuning
• [ ] Custom callbacks
• [ ] Multi-GPU training
• [ ] Custom training loops
• [ ] Training in the cloud
• [ ] TPU training

General

• [ ] Problem thinking
• [ ] Generative networks
• [ ] Experiment tracking
• [ ] Hyperparameter search
• [ ] Custom layers
• [ ] Advanced architectures

Advanced level

General

• [ ] Huge datasets
• [ ] Model deployment
• [ ] Multi-worker training
• [ ] Reinforcement learning
• [ ] Research
• [ ] Staying up-to-date