Datasets:

DDSC
/

partial-danish-gigaword-small-test-sample

dataset_info:
  features:
    - name: text
      dtype: string
    - name: source
      dtype: string
    - name: doc_id
      dtype: string
    - name: LICENSE
      dtype: string
    - name: uri
      dtype: string
    - name: date_built
      dtype: string
  splits:
    - name: train
      num_bytes: 23816547.04337273
      num_examples: 2411
  download_size: 11686492
  dataset_size: 23816547.04337273
language:
  - da
pretty_name: Danish Gigaword Test Sample

Dataset Card for "Danish Gigaword Test Sample"

This is a small sample of the dataset DDSC/partial-danish-gigaword-no-twitter. It is meant as a small dataset for testing code. It is constructed using the following code:

from datasets import concatenate_datasets, load_dataset

# download dataset from huggingface
dataset = load_dataset("DDSC/partial-danish-gigaword-no-twitter")

# All of the dataset is available in the train split - we can simply:
dataset = dataset["train"]

# downsample it to three domains
legal = dataset.filter(lambda x: x["source"] == "retsinformationdk")
news = dataset.filter(lambda x: x["source"] == "tv2r")
speech = dataset.filter(lambda x: x["source"] == "spont")

# downsample to 1000 samples
legal = legal.select(range(1000))
news = news.select(range(1000))

# combine the three domains
dataset = concatenate_datasets([legal, news, speech])

# upload to hub
dataset.push_to_hub("DDSC/partial-danish-gigaword-small-test-sample")