CommonCrawl Cleaned Dataset — 500GB

Description

Pre-processed and deduplicated CommonCrawl dataset. Filtered for English, removed boilerplate, PII-scrubbed. Ready for LLM pre-training.

Tags