data/adversarial_qa_droberta_question_context_answer/stats.train.json

{
  "examples": 10000,
  "inputs_max_tokens": 584,
  "inputs_tokens": 1917394,
  "targets_max_tokens": 212,
  "targets_tokens": 63677
}