---
dataset_info:
- config_name: tokenized_llama31
  features:
  - name: input_ids
    sequence: int32
  - name: labels
    sequence: int64
  - name: attention_mask
    sequence: int8
  splits:
  - name: train
    num_examples: 6699493
configs:
- config_name: tokenized_llama31
  data_files:
  - split: train
    path: tokenized_llama31/train-*
---