evaluate 0 0 0

mm-eval/MM-IFEval

published by @mm-eval

Alignment dataset tracked in the /datasets sub-catalog.

Lifetime downloads
462
Last 30 days
172
Likes
-
Size
77.5 MB
Created on HF
2026-05-26
Age
4 months ago

Tags

size_categories:n<1Kformat:parquetformat:optimized-parquetmodality:imagemodality:textlibrary:datasetslibrary:pandaslibrary:polarslibrary:mlcroissantregion:us

README current version from Hugging Face


dataset_info:
features:

  • name: id
    dtype: string
  • name: media
    list: image
  • name: messages
    dtype: string
  • name: answer
    dtype: string
  • name: question_type
    dtype: string
  • name: tag
    dtype: string
  • name: infer_type
    dtype: string
  • name: constraints
    dtype: string
  • name: num_constraints
    dtype: int64
  • name: del_cons
    dtype: string
  • name: source_id
    dtype: string
  • name: task
    dtype: string
    splits:
  • name: test
    num_bytes: 82055009
    num_examples: 687
    download_size: 81264429
    dataset_size: 82055009
    configs:
  • config_name: default
    data_files:
    • split: test
      path: data/test-*

README history 13 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. 2026-07-19metadata: migrate score_type -> score_pipeline (atomic stage contract; see mm...0aa82fa679 B
    Loading...
  2. 2026-07-08Add manifest (metadata.json)06eafd4679 B
    Loading...
  3. 2026-07-08Upload dataset78307e1679 B
    Loading...
  4. 2026-07-07metadata: 2026-07-07 revised taxonomy (task_type format vocabulary multiple_c...0da2f05340 B
    Loading...
  5. 2026-07-07Add manifest (metadata.json)d96fb4a340 B
    Loading...
  6. 2026-07-07Upload datasetc225f9e340 B
    Loading...
  7. 2026-07-07metadata: add score_type/score_params/score_protocol (official grading protoc...e2e2e16340 B
    Loading...
  8. 2026-06-01Tighten prompt_template_source.notes: explicit 5-tier classification per 2026...323edc3340 B
    Loading...
  9. 2026-05-30Refine prompt_template_source: lmms-eval/VLMEvalKit canonical citationse097609340 B
    Loading...
  10. 2026-05-30Populate prompt_template_source for all subsets per 2026-05-30 audite54e6fb340 B
    Loading...
  11. 2026-05-26Add v2 manifest (metadata.json)8a0bee3340 B
    Loading...
  12. 2026-05-26Upload dataset6e6642a340 B
    Loading...
  13. 2026-05-26initial commit35275650 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.

Related datasets same stage · evaluate

Read further

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("mm-eval/MM-IFEval")
Raw snapshot (Python)
from huggingface_hub import snapshot_download
snapshot_download(repo_id="mm-eval/MM-IFEval", repo_type="dataset")
git clone (requires git-lfs)
git clone https://huggingface.co/datasets/mm-eval/MM-IFEval
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