Hugging Face's logo

Datasets:
HPLT
/
2508-wds-evals

Modalities:
Tabular
Text
Formats:
parquet
Languages:
Spanish
French
Size:
10K - 100K
ArXiv:
Libraries:
Datasets
pandas
License:
Dataset card Data Studio Files
xet
 Community
2508-wds-evals / README.md
vmkhlv's picture
vmkhlv
Update README.md
03177c2 verified
preview code
|
raw
history blame contribute delete
4.33 kB
metadata 
license: apache-2.0
dataset_info:
  - config_name: fra_Latn
    features:
      - name: corpus
        dtype: string
      - name: category
        dtype: string
      - name: dataset
        dtype: string
      - name: task
        dtype: string
      - name: prompt
        dtype: string
      - name: model
        dtype: string
      - name: ckpt_num
        dtype: int64
      - name: score
        dtype: float64
    splits:
      - name: results
        num_bytes: 909987
        num_examples: 3784
    download_size: 33013
    dataset_size: 909987
  - config_name: spa_Latn
    features:
      - name: corpus
        dtype: string
      - name: category
        dtype: string
      - name: dataset
        dtype: string
      - name: task
        dtype: string
      - name: prompt
        dtype: string
      - name: model
        dtype: string
      - name: ckpt_num
        dtype: int64
      - name: score
        dtype: float64
    splits:
      - name: results
        num_bytes: 1511136
        num_examples: 6912
    download_size: 71232
    dataset_size: 1511136
configs:
  - config_name: fra_Latn
    data_files:
      - split: results
        path: fra_Latn/results-*
  - config_name: spa_Latn
    data_files:
      - split: results
        path: spa_Latn/results-*
language:
  - es
  - fr

HPLT 3.0: Details on WDS-based Sampling Evaluation Results

Dataset Description

This dataset contains fine-grained results from our HPLT 3.0 release evaluations comparing the new HPLT 3.0 corpora sampled using different Web Document Scorer (WDS) thresholds, focusing on Spanish and French. We compare three configurations: Top, Random, and Bottom. Random sampling represents the default approach, drawing uniformly on the full corpus, while Top and Bottom take advantage of the sorting by WDS levels and sequentially draw 100B training tokens from either end of the corpus. We pretrain 2.2B Llama-style decoder models on 100B tokens for each selected language and evaluate them using HPLT-E, a multilingual evaluation framework for comprehensive multi-prompt k-shot evaluation across 124 tasks and 500+ prompts in nine typologically diverse languages: Spanish (spa_Latn), French (fra_Latn), Czech (ces_Latn), Ukrainian (ukr_Cyrl), Finnish (fin_Latn), Catalan (cat_Latn), Galician (glg_Latn), Basque (eus_Latn), and Norwegian (Bokmål and Nynorsk; nor_Latn).

Please find more details in our paper and GitHub repository.

Uses

This dataset is intended for reproducibility and research purposes. Find an example on how to access the results:

from datasets import load_dataset

dataset = load_dataset("HPLT/2508-wds-evals", "spa_Latn", split="results").to_pandas()

Dataset Structure

Dataset Instances

Each dataset instance looks as follows:

{
  'corpus': 'Bottom',
  'category': 'Paraphrase detection',
  'dataset': 'paws_es',
  'task': 'paws_es_p2',
  'prompt': 'Oración 1: {{sentence1}}\nOración 2: {{sentence2}}\nPregunta: ¿Las oraciones 1 y 2 expresan el mismo significado? ¿Sí o no?\nRespuesta:',
  'model': '50B',
  'ckpt_num': 24000,
  'score': 45.35
}

Dataset Fields

  • corpus: corpus name (Top, Random, Bottom)
  • category: task category
  • dataset: evaluation dataset name
  • task: evaluation task (refers to a specific prompt)
  • prompt: prompt used for evaluation
  • model: number of pretraining tokens (B)
  • ckpt_num: number identifier for model
  • score: standard metric performance score

Cite Us 

@article{oepen2025hplt,
  title={HPLT\~{} 3.0: Very Large-Scale Multilingual Resources for LLM and MT. Mono-and Bi-lingual Data, Multilingual Evaluation, and Pre-Trained Models},
  author={Oepen, Stephan and Arefev, Nikolay and Aulamo, Mikko and Ba{\~n}{\'o}n, Marta and Buljan, Maja and Burchell, Laurie and Charpentier, Lucas and Chen, Pinzhen and Fedorova, Mariya and de Gibert, Ona and others},
  journal={arXiv preprint arXiv:2511.01066},
  year={2025}
}

Contact Us