Datasets
DATASETHugging Face
5/5

Tulu-3 SFT Mixture

Chewa(nya)TextInstruction-TuningEast Africa
View Source

Dataset Info

Language
Chewa (nya)
Modality
Text
Domain
General
Size
0
Institution
Allen Institute for AI (AI2)
Last Updated
3 Aug 2026

License & Access

SPDX identifierODC-BY

Licensed under ODC-BY. Review the source terms before use.

RedistributionYes

PII RISK

Low risk

Full Record

30 of 46 fields stated

Every column the registry tracks. Values marked derivedwere filled by a rule or read from the source's text rather than stated in its metadata; the rule or the quoted sentence is shown under the value.

Registry columns (23/27)

iso code
nya
Dataset name
Tulu-3 SFT Mixture
Modality
Text
Type
dataset
Link
https://huggingface.co/datasets/allenai/tulu-3-sft-mixture
NLP Task
Instruction-Tuning
Domain
General
Data Format
parquet
Size Metric
Sentences
Size Value
0
License
ODC-BY
Open Access
Open
Source Repository
Hugging Face
Year Published
2024
Creator/Author(s)
Allen Institute for AI (AI2)
Funder/Partner
Other
Dialect Coverage
Mixed
Gender Balance
Not stated
Speaker Count
Not stated
Annotation Type
Mixed
Status
Active
Has Datasheet
Yes
PII Risk
Low
Consent Verified
Not stated
Verified By
Not stated
Date Indexed
2026-08-03
Notes / Flags
Large general instruction-tuning mixture that includes some Chichewa-tagged data (via included multilingual subsets). Excluded as forks/derivatives/decontaminat…

Quality (DQI) (6/12)

Datasheet Fields Completed
Not stated
SPDX License ID
ODC-BY-1.0
Redistribution Allowed?
Yes
Content Hash
b14afda60f1bbebe55d5d2fa1e4df5042f97f8bederivedHugging Face card: commit sha
Submitted By
hf_data pipeline (automated)
Fetch Date
2026-08-03
Last Updated Date
2024-12-02derivedHugging Face card: lastModified
Annotation Guidelines Available?
Not stated
IAA Score
Not stated
Cleaning Notes
Not stated
Benchmark Exists?
Not stated
Benchmark Citation
Not stated

Representativeness (DRI) (1/7)

Gender Distribution
Not stated
Region Coverage
Not stated
Region Distribution
Not stated
Age Range Coverage
Not stated
Register
Not stated
Marginalized Group Note
Not stated
Demographic Fields Reported Count
0

Readiness Score Breakdown

5/5
Data Coverage100%
Model Availability100%
Benchmarks100%
Documentation85%
Recency100%

Representativeness

Gender Balance
Regional Distribution
Dialect Coverage
Inclusivity

Provenance

Source URL
https://huggingface.co/datasets/allenai/tulu-3-sft-mixture
Fetch Timestamp
3 Aug 2026
Content Hash
b14afda60f1bbebe55d5d2fa1e4df5042f97f8be

Audit Log

Record indexed from Hugging Face

3 Aug 2026

Source status: Active

3 Aug 2026

Other Copies

The same resource, also published at these locations. It is counted once.

dataset

tulu-3-sft-mixture

Data format: parquet

15 languages3 copiesChewa(nya)Textother
HUHugging Face
Open27 Oct 2026
dataset

tulu-3-sft-mixture

Data format: parquet

15 languages3 copiesChewa(nya)Textother
HUHugging Face
Open27 Oct 2026

Also Available In

This is the same multilingual corpus, sliced for other languages the platform indexes it under.

dataset

tulu-3-sft-mixture

Data format: parquet

15 languages3 copiesAmharic(amh)Textother
HUHugging Face
Open27 Oct 2026
dataset

tulu-3-sft-mixture

Data format: parquet

15 languages3 copiesMoroccan Arabic (Darija)(ary)Textother
HUHugging Face
Open27 Oct 2026
dataset

tulu-3-sft-mixture

Data format: parquet

15 languages3 copiesEgyptian Arabic(arz)Textother
HUHugging Face
Open27 Oct 2026
dataset

tulu-3-sft-mixture

Data format: parquet

15 languages3 copiesHausa(hau)Textother
HUHugging Face
Open27 Oct 2026
dataset

TULU-3 SFT Mixture

Large multilingual instruction tuning dataset with Igbo content

15 languages3 copiesIgbo(ibo)TextLM
HUHugging Face
Open30 Jun 2026
dataset

Tulu-3 SFT Mixture (AllenAI)

Tulu-3 SFT training mixture including Malagasy (mlg). Multiple forks and subsets excluded.

15 languages3 copiesMalagasy(mlg)Textother
HUHugging Face
Open17 Jul 2026
dataset

tulu-3-sft-mixture

Data format: parquet

15 languages3 copiesNorthern Sotho(nso)Textother
HUHugging Face
Open27 Oct 2026
dataset

tulu-3-sft-mixture

Data format: parquet

15 languages3 copiesShona(sna)Textother
HUHugging Face
Open27 Oct 2026
dataset

tulu-3-sft-mixture

Data format: parquet

15 languages3 copiesSomali(som)Textother
HUHugging Face
Open27 Oct 2026
dataset

tulu-3-sft-mixture

Data format: parquet

15 languages3 copiesSwahili(swa)Textother
HUHugging Face
Open27 Oct 2026
dataset

tulu-3-sft-mixture

Data format: parquet

15 languages3 copiesWolof(wol)Textother
HUHugging Face
Open27 Oct 2026
dataset

tulu-3-sft-mixture

Data format: parquet

15 languages3 copiesXhosa(xho)Textother
HUHugging Face
Open27 Oct 2026
dataset

tulu-3-sft-mixture

Data format: parquet

15 languages3 copiesYoruba(yor)Textother
HUHugging Face
Open27 Oct 2026
dataset

tulu-3-sft-mixture

Data format: parquet

15 languages3 copiesZulu(zul)Textother
HUHugging Face
Open27 Oct 2026

Related Datasets

dataset

FLEURS (Chichewa/Nyanja subset)

Canonical FLEURS release, which includes a Chichewa (nya) read-speech split. The raw HuggingFace nya-tagged query returned 11 additional personal re-uploads/…

27 languages15 copiesChewa(nya)SpeechASR
HUHugging Face
Open3 Aug 2026
dataset

FLORES-200 Evaluation Benchmark (Chichewa)

Consolidated FLORES-200/FLORES-101 family. The Lanfrica pull and HF search returned many repackagings of the same underlying FLORES data for nya: mteb/flores…

53 languages4 copiesChewa(nya)TextMT
HUHugging Face
Open3 Aug 2026
dataset

Aya Dataset

Canonical human-curated Aya instruction dataset, includes Chichewa. Excluded as forks/re-uploads: Felladrin/ChatML-aya_dataset (ChatML reformat), janavivekar…

15 languages4 copiesChewa(nya)TextInstruction-Tuning
HUHugging Face
Open3 Aug 2026