2026-10-11 18:01 UTC

Independent use will determine whether Hugging Face’s Stack v3, including its 114TB full corpus and deduplicated, quality-filtered, PII-redacted training split, becomes a standard open code-pretraining resource.

state: expiredheat: lowuncertainty: highnovelscott: noneopen-code-datasets code-training open-models hugging-faceHugging FaceAnton Lozhkov

What is this?

The Stack v3 is presented as a Hugging Face release for open code-model pretraining, comprising a complete 114TB research corpus and a ready-to-load training split described as near-deduplicated, quality-filtered, and PII-redacted. Its proposed significance is as shared infrastructure for open code models, but whether it becomes a standard resource depends on adoption and independent use. The supplied search snippets mostly document the original 3TB Stack and adjacent datasets rather than v3 itself, and they do not establish Anton Lozhkov’s role or independently verify the v3 release details.

Why it matters to Scott

No intersection found: neither Scott’s wikis nor the radar contain a position, project, or prior case connecting him to this dataset or its adoption as open code-training infrastructure.
queries asked of Scott's wikis
  • open code datasets as shared model infrastructure
  • code corpus deduplication quality filtering and PII removal
  • training-data openness versus open-weight models
  • dataset licensing and provenance for code models
  • standardization and network effects in AI datasets
  • code-pretraining data pipelines and evaluation

Measured heat

no measured readings yet — the hourly heat pass fills this in

How the heat travelled

no chain yet — the hourly chain pass fills this in

Evidence (3) — ⭐ canonical anchor

sourceobjectauthorscorecomments
🟠 redditHugging Face releases The Stack v3 – largest open code dataset yet
LocalLLaMA
Nunki0852483
🟧 echo.x ⭐Announced Stack v3 with a ready-to-load, near-deduplicated, quality-filtered and PII-redacted training dataset plus a complete 114TB corpus Anton Lozhkov——
🟧 hnAm I in the Stack? (AI training opt-out)lionkor10

Interpretation history

Decision trace