← Table of Contents · Page 849 of 868

APPENDIX H: RESUME PORTFOLIO BUILDING CAPSTONE PROJECTS

Part 39 of 56

STRETCH GOALS Add a sixth check that scores whether the summary is shorter than the original documents. Add a second scorer that uses a model as a judge, which means asking another model to grade an answer, and measure how often it agrees with your exact checks. Show a line chart of token use per case on the dashboard. Run each case five times with a real model and report the lowest, middle, and highest pass rates. Add a command that compares two result files and prints only the cases that changed.
Advertisement
Read Aloud:
0:00 / 0:00