lm-evaluation-harness/lm_eval/tasks/benchmarks/flan/flan_held_out.yaml

group: flan_held_out
task:
  # BBH
  - bbh_zeroshot
  - bbh_fewshot
  - bbh_cot_fewshot
  - bbh_cot_zeroshot
  # MMLU
  - mmlu
  - mmlu_flan_n_shot_generative
  - mmlu_flan_n_shot_loglikelihood
  - mmlu_flan_cot_zeroshot
  - mmlu_flan_cot_fewshot