179 questions
No questions match those filters.
← All questions
Training & Fine-tuningMedium· Alignment· SFT· DPO· RL
Walk me through how you would choose between SFT, DPO,...
This is one of the questions in the full AI/ML interview bank. Pro unlocks all 1789 questions; Premium includes the same bank plus the highest daily Practice limit.
See plansThe choice of training method depends on the type of feedback signal available. If you have expert demonstrations, start with SFT. If you have human pairwise preferences, move to DPO. If you have automatable correctness signals, use RLVR. You should layer these approaches only as far as your compute budget and signal quality allow, prioritizing SFT if you lack the necessary data to begin with.