Practice
Learn
← Quit
0 / 0 correct
LLMs & AI
Hard
1 of 20
How should you test changes to prompts or models in an LLM product?
A
Try a couple of examples by hand
B
Run a fixed evaluation suite of representative cases (with automatic and human scoring) before shipping, and monitor quality in production
C
Only check that the API returns 200
D
Measure tokenizer speed