google-research/deduplicate-text-datasets

• highabout#1

  Add a concise 'About' description

  原因：

  复制粘贴的修复

  复制

  Code and scripts for ExactSubstr deduplication of large text datasets (e.g., C4, LM1B) to improve language model training efficiency and reduce memorization. Includes Rust implementation and Python analysis tools.

  待办完成忽略
• mediumreadme#2

  Clarify the specific deduplication method offered in the README's opening

  原因：

  当前

  This repository contains code to deduplicate language model datasets as descrbed in the paper "Deduplicating Training Data Makes Language Models Better" by Katherine Lee, Daphne Ippolito, Andrew Nystrom, Chiyuan Zhang, Douglas Eck, Chris Callison-Burch and Nicholas Carlini. We release the ExactSubstr deduplication implementation (written in Rust) along with the scripts we used in the paper to perform ExactSubstr deduplication and inspect the results (written in Python).

  复制粘贴的修复

  复制

  This repository provides the **ExactSubstr deduplication implementation (written in Rust)** and associated Python scripts, as described in the paper "Deduplicating Training Data Makes Language Models Better". It enables efficient deduplication of large language model datasets to improve training and reduce memorization. We also release document clusters from NearDup deduplication on C4, RealNews, LM1B, and Wiki-4B-en.

  待办完成忽略

本次扫描解析到的品类 GEO 通道：google/gemini-2.5-flash, deepseek/deepseek-v4-flash

• 品类问题

  Seeking a method to efficiently deduplicate massive text corpora for training large language models.

  你：第 2 位
  AI 推荐顺序：

  1. datasketch
  2. deduplicate-text-datasets ← 你
  3. text_dedup
  4. Apache Spark
  5. dedupe
  查看 AI 完整回答
• 品类问题

  What are robust tools for identifying and removing duplicate content from very large text datasets?

  你：未被推荐
  AI 推荐顺序：

  1. Apache Spark
  2. Spark's MLlib
  3. MinHashLSH
  4. Dedupe.io (dedupeio/dedupe)
  5. Datasketch (ekzhu/datasketch)
  6. Dask (dask/dask)
  7. PySpark
  8. Elasticsearch (elastic/elasticsearch)
  9. Python
  10. fuzzywuzzy (seatgeek/fuzzywuzzy)
  11. difflib
  12. OpenRefine (OpenRefine/OpenRefine)

  AI 推荐了 12 个替代方案，却始终没点名 google-research/deduplicate-text-datasets。这就是要补上的差距。

  查看 AI 完整回答

• Metadata completeness

  fail

  建议：

• README presence

  pass

• Compared to common alternatives in this category, what is the core differentiator of google-research/deduplicate-text-datasets?

  pass
  ⚠AI 未点名 google-research/deduplicate-text-datasets —— 很可能在说另一个项目

  AI 的回答可能信誓旦旦却是错的。请按事实核对：技术栈、目标人群、差异化点是不是和你实际的对得上？

• If a team adopts google-research/deduplicate-text-datasets in production, what risks or prerequisites should they evaluate first?

  pass
  ✓AI 明确点名了 google-research/deduplicate-text-datasets

  AI 的回答可能信誓旦旦却是错的。请按事实核对：技术栈、目标人群、差异化点是不是和你实际的对得上？

• In one sentence, what problem does the repo google-research/deduplicate-text-datasets solve, and who is the primary audience?

  pass
  ⚠AI 未点名 google-research/deduplicate-text-datasets —— 很可能在说另一个项目

  AI 的回答可能信誓旦旦却是错的。请按事实核对：技术栈、目标人群、差异化点是不是和你实际的对得上？