1 phaseActive
Run a model on all 141 questions. Every question is one prompt of about 261,922 tokens.