Datasets:

aisingapore
/

SEA-PILE-v2

Tasks:

Text Generation

Modalities:

Formats:

Languages:

Size:

ArXiv:

Tags:

language-modeling

Libraries:

License:

Dataset card Data Studio Files Files and versions

SAnocha commited on Aug 12

Commit

77573cc

·

verified ·

1 Parent(s): 8140dd0

Update README.md

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -77,7 +77,7 @@ for language in languages:
 Despite our best efforts to filter out undesirable (i.e NSFW, toxic and biased) content and personally identifiable information (PII), there is a possibility that some documents containing harmful, toxic, or private content may still pass through our pipeline. We are committed to continuously improving our filtering processes to minimize these occurrences.
-*Note: The 9 rounded language rows in our dataset summary table total 122.0B tokens, which corresponds to the nominal 120B public release. This public dataset represents a curated, open-source subset of AI Singapore's broader internal pre-training pool (exceeding 1 Trillion tokens).*
 ## License

 Despite our best efforts to filter out undesirable (i.e NSFW, toxic and biased) content and personally identifiable information (PII), there is a possibility that some documents containing harmful, toxic, or private content may still pass through our pipeline. We are committed to continuously improving our filtering processes to minimize these occurrences.
+*Public Subset Distinction: This public repository contains the initial 120B/122B token open-source release. The expanded 1-trillion-token SEA-Pile v2 pool referenced in later SEA-LION model iterations (e.g., v4) represents an internal expansion containing additional regional web crawls and unreleased multi-source corpora.*
 ## License