GPT2-chitchat

--- dataset_info: features: - name: text dtype: string - name: input_ids sequence: int32 splits: - name: train num_bytes: 3121359 num_examples: 9800 - name: valid num_b