data/anli_MNLI_crowdsource_r1/stats.test.json

{
  "examples": 1000,
  "inputs_max_tokens": 224,
  "inputs_tokens": 128324,
  "targets_max_tokens": 5,
  "targets_tokens": 2665
}