Datasets:

Modalities:
Tabular
Text
Formats:
parquet
Languages:
Estonian
ArXiv:
Libraries:
Datasets
Dask
License:
adorkin commited on
Commit
1543d11
·
verified ·
1 Parent(s): 9f7889d

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +21 -1
README.md CHANGED
@@ -48,4 +48,24 @@ size_categories:
48
  - 1M<n<10M
49
  ---
50
 
51
- The Estonian subset of [HuggingFaceFW/fineweb-2](https://huggingface.co/datasets/HuggingFaceFW/fineweb-2)
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
48
  - 1M<n<10M
49
  ---
50
 
51
+ # FineWeb2-et
52
+
53
+ The Estonian subset of [HuggingFaceFW/fineweb-2](https://huggingface.co/datasets/HuggingFaceFW/fineweb-2) reuploaded for ease of access.
54
+
55
+ ## Licensing Information
56
+
57
+ The dataset is released under the **Open Data Commons Attribution License (ODC-By) v1.0** [license](https://opendatacommons.org/licenses/by/1-0/). The use of this dataset is also subject to [CommonCrawl's Terms of Use](https://commoncrawl.org/terms-of-use).
58
+
59
+ ## Citation Information
60
+
61
+ ```
62
+ @misc{penedo2025fineweb2pipelinescale,
63
+ title={FineWeb2: One Pipeline to Scale Them All -- Adapting Pre-Training Data Processing to Every Language},
64
+ author={Guilherme Penedo and Hynek Kydlíček and Vinko Sabolčec and Bettina Messmer and Negar Foroutan and Amir Hossein Kargaran and Colin Raffel and Martin Jaggi and Leandro Von Werra and Thomas Wolf},
65
+ year={2025},
66
+ eprint={2506.20920},
67
+ archivePrefix={arXiv},
68
+ primaryClass={cs.CL},
69
+ url={https://arxiv.org/abs/2506.20920},
70
+ }
71
+ ```