---
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
---