evaluate 0 0 0

TAUR-dev/IFEval__qwen__bon__single

published by @TAUR-dev

Alignment dataset tracked in the /datasets sub-catalog.

Lifetime downloads
273
Last 30 days
7
Likes
-
Size
3.1 MB
Created on HF
2025-07-09
Age
1 years ago

Tags

size_categories:n<1Kformat:parquetmodality:textlibrary:datasetslibrary:pandaslibrary:mlcroissantlibrary:polarsregion:us

README current version from Hugging Face


dataset_info:
features:

  • name: key
    dtype: int64
  • name: prompt
    dtype: string
  • name: instruction_id_list
    sequence: string
  • name: kwargs
    list:
    • name: capital_frequency
      dtype: float64
    • name: capital_relation
      dtype: string
    • name: end_phrase
      dtype: string
    • name: first_word
      dtype: 'null'
    • name: forbidden_words
      sequence: string
    • name: frequency
      dtype: float64
    • name: keyword
      dtype: string
    • name: keywords
      sequence: string
    • name: language
      dtype: string
    • name: let_frequency
      dtype: float64
    • name: let_relation
      dtype: string
    • name: letter
      dtype: string
    • name: nth_paragraph
      dtype: 'null'
    • name: num_bullets
      dtype: float64
    • name: num_highlights
      dtype: float64
    • name: num_paragraphs
      dtype: float64
    • name: num_placeholders
      dtype: float64
    • name: num_sections
      dtype: float64
    • name: num_sentences
      dtype: float64
    • name: num_words
      dtype: float64
    • name: postscript_marker
      dtype: string
    • name: prompt_to_repeat
      dtype: string
    • name: relation
      dtype: string
    • name: section_spliter
      dtype: string
  • name: prompt_chat_template
    list:
    • name: content
      dtype: string
    • name: role
      dtype: string
  • name: prompt__bon
    sequence: string
  • name: prompt_chat_template__bon__metadata
    struct:
    • name: api_url
      dtype: string
    • name: backend
      dtype: string
    • name: chat_template_applied
      dtype: bool
    • name: generation_params
      struct:
      • name: chat_template_applied
        dtype: bool
      • name: max_tokens
        dtype: int64
      • name: temperature
        dtype: float64
    • name: model_name
      dtype: string
    • name: prompt
      list:
      • name: content
        dtype: string
      • name: role
        dtype: string
  • name: prompt__bon__metadata
    struct:
    • name: backend
      dtype: string
    • name: model_name
      dtype: string
    • name: n_responses
      dtype: int64
  • name: response__single
    sequence: string
  • name: prompt__single__metadata
    struct:
    • name: api_url
      dtype: string
    • name: backend
      dtype: string
    • name: chat_template_applied
      dtype: bool
    • name: generation_params
      struct:
      • name: chat_template_applied
        dtype: bool
      • name: max_tokens
        dtype: int64
      • name: temperature
        dtype: float64
    • name: model_name
      dtype: string
    • name: prompt
      dtype: string
  • name: response__single__metadata
    struct:
    • name: backend
      dtype: string
    • name: model_name
      dtype: string
    • name: n_responses
      dtype: int64
      splits:
  • name: train
    num_bytes: 7142932
    num_examples: 100
    download_size: 3294498
    dataset_size: 7142932
    configs:
  • config_name: default
    data_files:
    • split: train
      path: data/train-*

README history 7 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. 2025-07-10Upload dataset9d68ffc3 KB
    Loading...
  2. 2025-07-10Upload dataset690b9ee3 KB
    Loading...
  3. 2025-07-10Upload datasetc839b843 KB
    Loading...
  4. 2025-07-10Upload dataset1f8fed23 KB
    Loading...
  5. 2025-07-10Upload dataset52689ea3 KB
    Loading...
  6. 2025-07-10Upload datasetc4708752.1 KB
    Loading...
  7. 2025-07-10initial commit0e49c790 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("TAUR-dev/IFEval__qwen__bon__single")
Raw snapshot (Python)
from huggingface_hub import snapshot_download
snapshot_download(repo_id="TAUR-dev/IFEval__qwen__bon__single", repo_type="dataset")
git clone (requires git-lfs)
git clone https://huggingface.co/datasets/TAUR-dev/IFEval__qwen__bon__single
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