Datasets:

pretraining
/

tiny-korean-100k

dataset_info:
  features:
    - name: text
      dtype: string
  splits:
    - name: train
      num_bytes: 520664221
      num_examples: 100000
  download_size: 309500222
  dataset_size: 520664221
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

import os
from datasets import load_dataset

# 1. Load your dataset
ds = load_dataset("minpeter/fineweb-2-edu-korean", split="train[:100_000]")

# 2. Drop all columns except "text"
other_cols = [c for c in ds.column_names if c != "text"]
ds = ds.remove_columns(other_cols)

# 3. Push to Hub
ds.push_to_hub(
    repo_id="pretraining/tiny-korean-100k",
    private=False,
    split="train",
    token=os.environ["HF_TOKEN"],
    commit_message="Upload tiny Korean dataset for pretraining (text only)",
)

print("Dataset uploaded successfully!")