Putting it together: designing a model rather than reusing one, deciding when to train from scratch, and making a model’s decisions explainable. Part of the ML Engineer track.

  1. Designing an Architecture — [Advanced] — choosing and composing an architecture for a problem
  2. Training From Scratch vs Fine-Tuning — [Advanced] — when each is worth it
  3. Fine-Tuning — [Advanced] — specializing a pretrained model, LoRA/PEFT
  4. Explainable AI — [Advanced] — SHAP, LIME, and why interpretability matters for trust and debugging
↑ Back to 02 — ML Engineer