misc 0 0 0

zhyfei314/Mynary-Offline-Dictionary

published by @zhyfei314

Alignment dataset tracked in the /datasets sub-catalog.

Lifetime downloads
1K
Last 30 days
1K
Likes
-
Size
901.0 MB
Created on HF
2026-09-22
Age
29 days ago

Description

Mynary Offline Dictionary

This dataset contains compact, downloadable offline dictionary packs for the

Mynary Dictionary Obsidian plugin.

The packs are generated from Wiktionary extracts published by

Kaikki.org. They are intended for local lookup and

language-learning workflows. The dataset is not a language model and is not

intended for training or benchmarking without additional preprocessing.

Repository layout

Each pack is stored under release/<pack-id>/ and… See the full description on the dataset page: https://huggingface.co/datasets/zhyfei314/Mynary-Offline-Dictionary.

Tags

task_categories:text-retrievallanguage:arlanguage:bglanguage:cslanguage:dalanguage:delanguage:enlanguage:eslanguage:etlanguage:filanguage:frlanguage:helanguage:hilanguage:hulanguage:idlanguage:itlanguage:jalanguage:kolanguage:ltlanguage:lvlanguage:nllanguage:pllanguage:ptlanguage:rolanguage:rulanguage:sklanguage:sllanguage:svlanguage:thlanguage:trlanguage:uklanguage:vilanguage:zhlicense:otherregion:usdictionarywiktionaryofflinemultilingualbilingual

README current version from Hugging Face


license: other
license_name: per-pack-source-licenses
license_link: https://github.com/zhyfei314/Mynary/blob/master/LICENSE-CC-BY-SA-4.0.txt
task_categories:

  • text-retrieval
    language:
  • ar
  • bg
  • cs
  • da
  • de
  • en
  • es
  • et
  • fi
  • fr
  • he
  • hi
  • hu
  • id
  • it
  • ja
  • ko
  • lt
  • lv
  • nl
  • pl
  • pt
  • ro
  • ru
  • sk
  • sl
  • sv
  • th
  • tr
  • uk
  • vi
  • zh
    pretty_name: Mynary Offline Dictionary
    tags:
  • dictionary
  • wiktionary
  • offline
  • multilingual
  • bilingual
  • obsidian

Mynary Offline Dictionary

This dataset contains compact, downloadable offline dictionary packs for the
Mynary Dictionary Obsidian plugin.

The packs are generated from Wiktionary extracts published by
Kaikki.org. They are intended for local lookup and
language-learning workflows. The dataset is not a language model and is not
intended for training or benchmarking without additional preprocessing.

Repository layout

Each pack is stored under release/<pack-id>/ and contains only the files
needed by the plugin:

release/<pack-id>/
├── manifest.json
├── index.json
└── entries.jsonl.gz

The current catalog contains 50 packs covering core dictionaries and selected
bilingual dictionaries. Large compressed entry files are kept in this Dataset
repository; the small catalog used by the plugin is maintained separately in
the GitHub repository under dictionary-catalog/catalog.json.

Pack format

Every manifest.json uses the mynary-pack-v1 schema. Core and bilingual
packs share the same fields:

{
  "format": "mynary-pack-v1",
  "id": "en-core",
  "language": "en",
  "name": "English Core Dictionary",
  "version": "2026-09-09",
  "entryCount": 1355096,
  "indexFile": "index.json",
  "entriesFile": "entries.jsonl.gz",
  "compressed": true,
  "source": "https://kaikki.org/",
  "license": "CC BY-SA 4.0 / GFDL",
  "indexSha256": "...",
  "entriesSha256": "...",
  "kind": "core"
}

For a bilingual pack, kind is bilingual and targetLanguage identifies
the translation direction. For example, an English-Vietnamese pack has
language: "en" and targetLanguage: "vi".

index.json contains a compact list of words and byte offsets. Each line in
the UTF-8 entries.jsonl.gz file is one JSON entry. The SHA-256 values in the
manifest are calculated over the exact downloadable bytes, including the gzip
file itself. Consumers should verify both checksums before installing a pack.

Use with Mynary

  1. Open Settings → Mynary Dictionary.
  2. Enable Use offline dictionary packs.
  3. Open Offline dictionary packs and select Choose language.
  4. Select a pack and choose Install.

Mynary downloads the selected manifest, index, and compressed entries file,
verifies the checksums, and stores the pack inside the vault. If an installed
pack does not contain a requested word, Mynary can fall back to Wiktionary
when online.

Data and licensing

The dictionary content is derived from Wiktionary and follows the applicable
source licensing terms. License requirements can vary by source snapshot and
pack; always check the license field in the pack's manifest.json before
redistributing a pack.

The repository includes reference copies of the relevant license texts:

When redistributing these files, retain attribution to Wiktionary and the
relevant source providers, preserve license notices, and provide the required
license text or a link to it. Do not assume that the MIT license of the Mynary
plugin applies to the dictionary data.

Citation and attribution

Suggested attribution:

Dictionary data derived from Wiktionary, processed from Kaikki.org extracts,
and packaged for Mynary Offline Dictionary.

Useful links:

Limitations

  • Coverage and quality vary by language and Wiktionary entry.
  • This dataset contains dictionary entries, not audio files or source HTML.
  • Pack versions correspond to the source snapshot recorded in each manifest.
  • The compressed JSONL files are optimized for the Mynary client and may need
    conversion before use in other tools.

README history 6 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. 2026-09-23Upload catalog.json with huggingface_huba7659f54.5 KB
    Loading...
  2. 2026-09-22Upload catalog.json with huggingface_huba1ddb4f4.5 KB
    Loading...
  3. 2026-09-22Upload README.mdbb273024.5 KB
    Loading...
  4. 2026-09-22Upload 3 filesd61792c0 B
    Loading...
  5. 2026-09-22Publish Mynary offline dictionary packsb0d31890 B
    Loading...
  6. 2026-09-22initial commitd77af410 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 · misc

Read further

Metadata

License
other

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("zhyfei314/Mynary-Offline-Dictionary")
Raw snapshot (Python)
from huggingface_hub import snapshot_download
snapshot_download(repo_id="zhyfei314/Mynary-Offline-Dictionary", repo_type="dataset")
git clone (requires git-lfs)
git clone https://huggingface.co/datasets/zhyfei314/Mynary-Offline-Dictionary
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