Dataset Viewer
Auto-converted to Parquet Duplicate
index
int64
0
1.12M
text
large_stringlengths
2.64k
150k
len
float64
2.64k
150k
lines
float64
0
4.11k
546,477
The name Maccabee, probably meaning “hammer,” is actually applied in the Books of Maccabees to only one man, Judas, third son of the priest Mattathias and first leader of the revolt against the Seleucid kings who persecuted the Jews (1 Mc 2:4, 66; 2 Mc 8:5, 16; 10:1, 16). Traditionally the name has come to be extended ...
149,184
586
837,190
"| Greeley Co. NEGenWeb|\n| A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | (...TRUNCATED)
145,101
340
723,884
"1:1 FORASMUCH as many have taken in hand to set forth in order a narration of the things that have (...TRUNCATED)
138,354
1,150
462,610
"THE LIFE OF\nGENERAL HUGH MERCER\nGENERAL HUGH MERCER\nWith brief sketches of General George Wash\n(...TRUNCATED)
137,641
3,594
136,238
"The History of Last Names •\nSurnames A-Z •\nWhen communities consisted of just a few people, s(...TRUNCATED)
134,974
2,252
984,529
"Inventions Made for Hire\nDespite the continued reliance on the rhetorical device that modern inven(...TRUNCATED)
129,402
285
352,050
"Liz Augereau, Sally Lloyd, Caroline Glass, Barry Johnston and Hugh Casement have spent enormous eff(...TRUNCATED)
127,145
492
898,815
"janilye on Family Tree Circles\nJournals and Posts\nThe Orient steamship John Elder, A. J. Cooper c(...TRUNCATED)
126,171
746
1,043,832
"The Holy Bible\nOld and New Testament\nAbout | PDFs | Mobile formats | Word formats | Other languag(...TRUNCATED)
124,917
119
963,291
"The Apocalypse, or Revelation to John, the last book of the Bible, is one of the most difficult to (...TRUNCATED)
123,792
771
End of preview. Expand in Data Studio

Fineweb-Tiny

Dataset Description

Fineweb-Tiny is a highly curated, premium subset extracted from nampdn-ai/mini-fineweb.

How "The Best" Was Determined

This dataset was created programmatically by streaming the original dataset and sorting chunks based on a rigorous quality scoring algorithm. The heuristic heavily favors:

  1. High language_score (if provided by the upstream extraction).
  2. Optimal document length (penalizing abnormally short snippets and excessively long, unformatted dumps).
  3. Strong structural coherence suitable for pre-training Small Language Models (SLMs) and LLMs.

Fineweb-Tiny contains the absolute BEST 72.9 Gigabytes (compressed Parquet) of the original source. It drops the bottom 50% of lower-quality data from the stream to ensure dense, high-utility text.

License

This dataset is released under the Open Data Commons Attribution License (ODC-By) v1.0 to perfectly match the upstream Fineweb mini licensing constraints.

Downloads last month
116