1 phaseActive
Run a model on all 236 questions. Every question is one prompt of about 524,149 tokens.