1 phaseActive
Run a model on all 310 questions. Every question is one prompt of about 1,048,384 tokens.