Datasets:

root-sys-n
/

mima-pona-dataset

text stringlengths 0 592
sina sona ala sona e ni: jan Ton li lili la ona li lon ma Oselija?
ona mute li wile lape.
mi wile e ni: mi jan ante.
mi mama mi ala.
poki li jaki a!
If I'd known her address, I would've written her a letter.
I couldn't help but fall in love with you.
sina mute li sona ala sona e jan Ton?
toki Losi li kiwen tawa kama sona.
This is why I was late for school.
How did you know that I was lying?
If everything happens for a reason, does that reason have a reason to be?
o pana e poki moku mi tawa mi.
contesto, argomento, salienza, pertinente, attuale, pertinenza, essere rilevante
Translated, ""Tokyo"" means ""eastern capital.
ona li tawa lon kon kepeken soweli tawa waso.
What?
We both kissed Tom yesterday.
a strong person
tenpo pini la ijo pi kalama musi li pona lili kin.
toki. ni la tomo telo li lon anu seme?
Tom replaced John as captain.
ona li pali e ni lon tenpo ni.
The strong must help the weak.
後ろ、後部、背後、背中
* sina ken ala ken kama tawa tomo mi
outil, instrument, équipement, machine, appareil
The cheese is yellow.
I went on foot.
mije olin ona en soweli Manki li sama lukin.
sitelen e toki
soweli o, kama tan ni a!
She does know it.
kin la
All birds come from inside an egg.
What is the number?
Why do you think that's a good idea?
Can I open the window?
We were late because of you.
kon pi telo kasi seli ni li pona.
pipi li wile pali e len alasa.
Who was on the phone this morning?
Can somebody get that?
There's a rock in my shoe.
medicine, medical
My luggage didn't arrive. What happened?
The canal connects the Atlantic and Pacific Oceans.
I don't want to stop.
ma tomo Pasen
mi sona e ni: sina kama sona mute.
pona sina la sina ken ala ken kepeken sin e ilo toki lon tenpo kama?
You know where it is, don't you?
sona mi la, nimi ni li lon ala.
mi kute e kalama ala.
Tom wanted a fresh start.
mi jo e tenpo kulupu mute lon poka pi jan Ton.
sina sona ala sona e ni: sina pakala e nasin lawa?
o kama jo e kon.
soweli nena lili li suwi.
[forma alternativă a lui ale)
empezar, comenzar, iniciar; abrir; encender, prender
mi wile unpa monsi e sina.
jan Jani li tawa tomo seli.
tenpo weka pali li pona tawa sina anu seme?
drøm
VERBO scii, koni, esti lerta pri, esti saĝa pri, havi informojn pri
ANTAŬVERBO scipovi
Keep your eyes closed until I say that you should open them.
There are a lot of stairs in this apartment building.
I don't dance.
Cool Autumn arrived.
mangiare, bere, consumare, ingoiare, ingerire
ona li pilin jaki tan moku.
He doesn't want to see me.
His book is full of utopias.
sina kule ala kule ante e linja sina?
sina lukin sama ni: sina kama tan supa lape.
Poverty leads to crime.
mi sona pi mute ali.
pilin ike li sama ijo suli. pilin pona li lili.
I would not ask if I knew the answer.
Nigeria
(आश्चर्य और चौंकने का प्रतीक)
Tom called me a coward.
They are all dead.
That's like a kick in the nuts.
patr(in)o, praulo; kreinto, iniciatinto, fondinto; prizorganto, bontenanto
jan lili li musi lon ma pi kasi kule.
veza, poveznica, zajednička karakteristika; križ, čvor, spojnica, sjecište; preklopiti, saviti; prekoračiti, premašiti do točke
(کسی اور کے ہوالے کے شروعات کی علامت)
What have you decided?
The strings are tied together.
Where are we going tonight?
o toki tawa ona. o toki e pilin sina tawa ona.
jan pona pi jan sewi li pilin e ni: jan sewi li lon ala.
mi kama sona e toki Kalike.
Neither Tom nor Mary goes to the store.
sina pali lon tomo seme?
They need daily care.
tenpo ale la kulupu jan li ante e ona.

End of preview. Expand in Data Studio

YAML Metadata Warning:The task_categories "language-modeling" is not in the official list: text-classification, token-classification, table-question-answering, question-answering, zero-shot-classification, translation, summarization, feature-extraction, text-generation, fill-mask, sentence-similarity, text-to-speech, text-to-audio, automatic-speech-recognition, audio-to-audio, audio-classification, audio-text-to-text, voice-activity-detection, depth-estimation, image-classification, object-detection, image-segmentation, text-to-image, image-to-text, image-to-image, image-to-video, unconditional-image-generation, video-classification, reinforcement-learning, robotics, tabular-classification, tabular-regression, tabular-to-text, table-to-text, multiple-choice, text-ranking, text-retrieval, time-series-forecasting, text-to-video, image-text-to-text, image-text-to-image, image-text-to-video, visual-question-answering, document-question-answering, zero-shot-image-classification, graph-ml, mask-generation, zero-shot-object-detection, text-to-3d, image-to-3d, image-feature-extraction, video-text-to-text, keypoint-detection, visual-document-retrieval, any-to-any, video-to-video, other

Mima Pona — Toki Pona Dataset

Clean Toki Pona language dataset for training small language models (≤5M parameters). All sentences are verified to contain only Toki Pona vocabulary — no English, Polish, Hindi, or other language contamination.

Dataset Structure

The dataset is provided as TSV files with two columns:

Column	Description
`source`	Data source label (e.g., `adam-sentences`, `jasima`, `lipu-sona`, `wikipedia`, `adam-documents`, `adam-chapters`, `lipu-tenpo`)
`sentence`	Clean Toki Pona sentence

Example rows

source	sentence
adam-sentences	soweli Manki ni li pi kulupu seme?
adam-sentences	lipu pi kasi suli li kama loje.
adam-sentences	mi pana e pona tawa jan Ton.
jasima	dobry, pozytywny, użyteczny; miły, pokojny; prosty
jasima	mi wile e ni: mi jan ante.

Note: The jasima source contains dictionary definitions in Toki Pona (def.tok field from lipu Linku), which may look like glosses rather than full sentences. These are still valid Toki Pona text.

Splits

Split	Sentences	Characters	Average Length
Train	31,158	1,280,634	41.1 chars
Validation	3,463	145,662	42.1 chars
Total	34,621	1,426,298	41.3 chars

Source Distribution

Source	Sentences	Description
`adam-sentences`	33,297	adam-mcdaniel/toki-pona-dataset — sentence corpus
`lipu-sona`	555	pona-la/lipu-sona — Toki Pona course lessons
`jasima`	243	lipu-linku/jasima — dictionary definitions in Toki Pona (`def.tok`)
`adam-documents`	397	adam-mcdaniel — document corpus
`adam-chapters`	42	adam-mcdaniel — chapter texts
`wikipedia`	80	Toki Pona Wikipedia extracts
`lipu-tenpo`	7	lipu tenpo homepage
Total	34,621

Data Cleaning

All sentences were filtered to contain only Toki Pona vocabulary (~137 core words from lipu Linku). Non-Toki-Pona content (English translations, Polish definitions, Hindi glosses, etc.) was removed. Each sentence contains at least 2 Toki Pona core words or has a TP word ratio ≥30%.

Intended Use

Training small language models (1M–5M parameters) on Toki Pona — a constructed language with ~120–140 core words. Suitable for:

Small Transformer language modeling
Text generation in Toki Pona
Tokenizer training (word-level vocab ~200 tokens)

Not Suitable For

English or other natural language tasks
Tasks requiring multi-lingual content (dataset is Toki Pona-only)

Citation

If you use this dataset, please cite:

@dataset{mima-pona-dataset,
  author = {root-sys-n},
  title = {Mima Pona — Toki Pona Language Dataset},
  year = {2026},
  url = {https://huggingface.co/datasets/root-sys-n/mima-pona-dataset}
}

Contact

Questions or issues: open an issue on the Hugging Face dataset page.

Downloads last month: 41