README.md

metadata

language:
  - pt
license: other
size_categories:
  - 1M<n<10M
task_categories:
  - text-generation
pretty_name: anvisa_instruct_tokenized
dataset_info:
  features:
    - name: input_ids
      sequence: int32
    - name: attention_mask
      sequence: int8
    - name: labels
      sequence: int64
  splits:
    - name: train
      num_bytes: 80805366840
      num_examples: 3033690
    - name: test
      num_bytes: 799080000
      num_examples: 30000
  download_size: 24242493429
  dataset_size: 81604446840
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
tags:
  - portuguese
  - language-modeling