Compare

Same-epoch comparison of training runs across wandb, neptune, tensorboard, or mlflow. Aligns runs at the student's current step (never current-vs-final-of-baseline) and separates proxy metrics from downstream targets. Use when the user asks to compare runs, check if a run is improving, track lag against a baseline, rank experiments, or evaluate run-vs-run performance.

fcakyon Updated 1.2k repo stars
Find papers, reproduce, design, launch, compare and publish, with compare looping back to design for the next experiment

File contents

fcakyon/phd-skills commit aa713ee05d

Frequently asked questions

npx get-skill add fcakyon/compare