Datasets:

AletheiaAING
/

Data-Quality

Tasks:

Text Classification

Languages:

Size:

Tags:

machine-generated

african-languages

synthetic-degradation

License:

Dataset card Files Files and versions

691 MB

Ctrl+K

Ctrl+K

1 contributor

History: 62 commits

BeardedMonster's picture

CPU docs @ 2026-10-21 02:06 UTC

e86753f verified 2 minutes ago

1b_inlang
in_lang_1b: 563,485 rows, 69 languages about 10 hours ago
270m
270m: 3,742 rows from google/gemma-3-270m 3 days ago
270m_in_lang
box C in_lang: 1,641 rows, 68 languages, orthography scored 2 days ago
270m_inlang_b
in_lang_270m: 45,560 rows, 71 languages about 20 hours ago
270m_inlang_b2
in_lang_270m: 18,268 rows, 67 languages about 12 hours ago
270m_inlang_full
in_lang_270m: 57,465 rows, 71 languages 1 day ago
270m_pool_b
box C pool_b: 1,247 rows, 20 languages, orthography scored 2 days ago
diacritics_sft
diacritic restoration SFT: 135,116 rows, 39 languages, 40/60 full/partial 1 day ago
diacritics_sft_v2
diacritics_sft_v2: 163,480 rows, 40 languages, +AfriADR train about 22 hours ago
gemma-pretrain_monolingual
CPU docs @ 2026-10-21 02:06 UTC 2 minutes ago
.gitattributes

2.5 kB
initial commit 3 days ago
README.md
7.86 kB
index: add 270m_inlang_b, record the zero A/B prompt overlap about 20 hours ago
quality.parquet

211 MB
xet

493,831 rows, 71 languages, synthetic degradation + LLM-judge prompts 3 days ago
stats.json

1.31 kB
493,831 rows, 71 languages, synthetic degradation + LLM-judge prompts 3 days ago