data/adversarial_qa_droberta_answer_the_following_q/stats.train.json

{
  "examples": 10000,
  "inputs_max_tokens": 599,
  "inputs_tokens": 2051394,
  "targets_max_tokens": 212,
  "targets_tokens": 63677
}