evaluate 0 0 0

shrugging-shoulders/inverse_ifeval_multilang

published by @shrugging-shoulders

Alignment dataset tracked in the /datasets sub-catalog.

Lifetime downloads
87
Last 30 days
34
Likes
-
Size
12.4 MB
Created on HF
2026-09-02
Age
1 months ago

Description

A dataset for improving instruction following capabilities, based on inverse ifeval dataset.

Included languages:

English

French

German

Spanish

Italian

Portuguese

Russian

Chinese

Japanese

Does have exact same sample structures as inverse ifeval dataset (or very similar), but has slightly different constraints and has multiple languages.

Language distribution:

1:1 - every language is evenly present

Tags

task_categories:text-generationlanguage:enlanguage:frlanguage:delanguage:eslanguage:itlanguage:ptlanguage:rulanguage:zhlanguage:jalicense:apache-2.0size_categories:1K<n<10Kformat:parquetmodality:textlibrary:datasetslibrary:pandaslibrary:polarslibrary:mlcroissantregion:us

README current version from Hugging Face


dataset_info:
features:

  • name: messages
    list:
    • name: role
      dtype: string
    • name: content
      dtype: string
  • name: language
    dtype: string
  • name: capability_target
    dtype: string
    splits:
  • name: train
    num_bytes: 13366927
    num_examples: 8411
    download_size: 13021295
    dataset_size: 13366927
    configs:
  • config_name: default
    data_files:
    • split: train
      path: data/train-*
      license: apache-2.0
      language:
  • en
  • fr
  • de
  • es
  • it
  • pt
  • ru
  • zh
  • ja
    task_categories:
  • text-generation
    size_categories:
  • 1K<n<10K

A dataset for improving instruction following capabilities, based on inverse ifeval dataset.

Included languages:

  • English
  • French
  • German
  • Spanish
  • Italian
  • Portuguese
  • Russian
  • Chinese
  • Japanese

Does have exact same sample structures as inverse ifeval dataset (or very similar), but has slightly different constraints and has multiple languages.

Language distribution:
1:1 - every language is evenly present

README history 4 revisions

Every night we snapshot the README of every dataset in the catalog. When the SHA changes we archive the new version and diff it against the last. This is the evolving thought record of the alignment-data field: what the author decided to say about the corpus, and how that framing shifted over time.

  1. 2026-09-02Update README.mdf942766991 B
    Loading...
  2. 2026-09-02Update README.md95d8cef853 B
    Loading...
  3. 2026-09-02Upload dataset40e38e4433 B
    Loading...
  4. 2026-09-02initial commit2ad20fe0 B
    Loading...

Used by abliterated models none yet

No abliterated models in our catalog have declared this dataset in their YAML metadata yet. This can mean the dataset is used but not declared, is used outside abliteration workflows, or is new. The link will populate automatically as models are indexed.

Community discussions 1 thread · 1 comments

Threads posted by the community on Hugging Face: bug reports, questions about usage, pull requests on the README, license clarifications. Click any thread to expand and read the full conversation inline. Newest first.

  1. 2026-09-03[bot] Conversion to Parquetclosed1 💬#1
    Loading...

Related datasets same stage · evaluate

Read further

Metadata

License
apache-2.0

Repository

Download the dataset

0

Copy any of these snippets into your notebook or terminal. All three fetch directly from Hugging Face using your own credentials.

Recommended: datasets library (Python)
from datasets import load_dataset
ds = load_dataset("shrugging-shoulders/inverse_ifeval_multilang")
Raw snapshot (Python)
from huggingface_hub import snapshot_download
snapshot_download(repo_id="shrugging-shoulders/inverse_ifeval_multilang", repo_type="dataset")
git clone (requires git-lfs)
git clone https://huggingface.co/datasets/shrugging-shoulders/inverse_ifeval_multilang
Catalog is the map. Apps are the tools.

Run models on your own machine, not in the cloud.

Every model page has an "Open in Abliteration" button that hands the model directly to the first-party desktop client, at the quantization your rig can actually run. No API keys, no subscription, no prompt leakage.

Open in Abliteration