da-instruct-dynaword-hq

Danish instruction fine-tuning dataset generated via backtranslation from danish-foundation-models/danish-dynaword, filtered to high-quality samples using danish-foundation-models/dynaword-annotations.

All 40 DynaWord subsets are included — both contemporary and historical Danish. See oliverkinch/da-instruct-dynaword-contemporary-hq for a version restricted to contemporary Danish sources.

Dataset description

Each row is a (prompt, target) pair where:

• target is a passage of authentic Danish text drawn from a curated subset of DynaWord
• prompt is a realistic Danish user instruction that would plausibly elicit that text from a language model

Prompts were generated by Qwen/Qwen3.5-397B-A17B using a backtranslation approach: given a passage and its source context, the model is asked to write the user message that would cause a chatbot to produce a similar text. Domain-specific prompt templates (government, literary, academic, tax guidance, news, speech) ensure that the generated instructions match the register and genre of the source text.

Construction

1 — Quality filtering with dynaword-annotations

The dynaword-annotations dataset provides per-sample quality annotations for the full 5.66M-sample DynaWord corpus. The following filter was applied:

Criterion	Allowed values
content_integrity	complete
content_quality	excellent
content_safety	safe
pii_presence	no_pii

9,200 IDs were randomly sampled without replacement from the pool of qualifying IDs across all 40 dynaword subsets.

2 — Passage extraction

Articles shorter than 400 characters were discarded. Longer articles were chunked into passages of up to 2,500 characters using a hybrid scoring approach (paragraph boundary detection with quality scoring). A maximum of 2 passages per article was kept.

Passage-level quality filters reject:

• OCR noise (inter-word double spaces, ligature splits, high mixed-alphanumeric token ratio)
• Repetitive content (unique word ratio < 28% or word run ≥ 6)
• Document colophons and mastheads (phone/email/ISSN markers in first 300 chars)
• YAML/markdown frontmatter

3 — Prompt generation

For each passage, Qwen3.5-397B-A17B was prompted with the passage text and source context metadata to generate a Danish user instruction. Six domain-specific system prompts were used:

• government — administrative and public information requests
• literary — prose, fiction, letters and historical texts
• academic — scholarly articles and analytical texts
• tax_guidance — practical citizen queries about tax rules
• speech — parliamentary speeches and public addresses
• news — news articles and encyclopedic texts (default)

Generated prompts were validated with rule-based filters: language detection (must be Danish), user-request detection (must include a question or imperative verb), shouting token check, roleplay pattern rejection, and prompt-target content overlap check.

4 — Results

10,660 instruction pairs generated from 9,200 randomly sampled source IDs across 40 dynaword subsets.

Source coverage

Subset	Domain	Source type	Pairs
retsinformationdk	legal	government	4,729
enevaeldens_nyheder	historical news	news	1,635
cellar	EU documents	government	1,280
wiki	encyclopedia	news	562
ai-aktindsigt	government	government	473
skat	tax	tax_guidance	327
health_hovedstaden	health	government	304
Domsdatabasen	legal	government	267
municipality_meetings	government	government	263
ncc_maalfrid	web/government	government	125
kb_historical_letters	historical letters	literary	117
tidsskrift-dk	journalism/academic	academic	95
kb_administrative_publication	administrative	government	68
miljoeportalen	environment	government	65
retspraksis	legal	government	60
ep	EU parliament	speech	50
ncc_books	books	literary	42
danske-taler	speeches	speech	27
opensubtitles	film/TV subtitles	literary	27
eur-lex-sum-da	EU law	government	25
fm-udgivelser	government	government	22
wikisource	books	literary	19
adl	literature	literary	16
memo	books	literary	13
ncc_parliament	parliament	speech	13
grundtvig	literary	literary	11
ft	parliament	speech	8
wikibooks	educational	news	8
nordjyllandnews	news	news	4
gutenberg	books	literary	2
naat	web	speech	2
wiki_misc	wiki comments	news	1

Schema

{
    "id": "backtranslation_passages_dynaword-<hash>",
    "prompt": "Danish user instruction (generated)",
    "target": "Danish text passage (from source corpus)",
    "meta": {
        "passage_idx": 0,                # passage index within the source article
        "source_config_name": "skat",    # HuggingFace config name
        "source_dataset": "danish-foundation-models/danish-dynaword",
        "source_id": "skat::SKM...",     # original article ID
        "source_key": "skat",
        "source_name": "skat",           # dynaword subset name
        "source_record_index": 42,
        "source_split": "train",
        "source_type": "tax_guidance",   # prompt template used
        "target_chars": 1842
    },
    "sources": [
        {"config_name": "skat", "dataset": "danish-foundation-models/danish-dynaword",
         "row_id": "skat::SKM...", "split": "train"}
    ]
}

Usage

from datasets import load_dataset

ds = load_dataset("oliverkinch/da-instruct-dynaword-hq")
# ds["train"]

# Format for SFT
def format_example(row):
    return {
        "messages": [
            {"role": "user",      "content": row["prompt"]},
            {"role": "assistant", "content": row["target"]},
        ]
    }

Licence

This dataset inherits the licences of the constituent DynaWord subsets. Subsets carry a range of open licences (CC0, CC-BY, CC-BY-SA, government open data). See the DynaWord dataset card for per-subset licence details.

The generated prompt column was produced with Qwen/Qwen3.5-397B-A17B and is released under CC-BY 4.0.

Related datasets

Dataset	Subsets	Quality filter	Pairs
oliverkinch/da-instruct-dynaword-contemporary	contemporary only	none	8,271
oliverkinch/da-instruct-dynaword-contemporary-hq	contemporary only	dynaword-annotations	11,303
oliverkinch/da-instruct-dynaword-hq	all 40 subsets	dynaword-annotations	10,660
oliverkinch/da-instruct-dynaword	all 40 subsets	dynaword-annotations	39,847

• danish-foundation-models/danish-dynaword — source corpus
• danish-foundation-models/dynaword-annotations — quality annotations used for filtering