---
license: apache-2.0
task_categories:
  - text-generation
language:
  - en
configs:
  - config_name: default
    data_files:
      - split: train
        path:
                - shard_00000.parquet
                - shard_00001.parquet
                - shard_00002.parquet
                - shard_00003.parquet
                - shard_00004.parquet
---

# Dataset

Mixed pretraining corpus - 1,000,000 rows, 5 shards.

Deduped (exact=False, fuzzy=False).