Summary of Datasets For Large Language Models: a Comprehensive Survey, by Yang Liu et al.
Datasets for Large Language Models: A Comprehensive Survey
by Yang Liu, Jiahuan Cao, Chongyu Liu, Kai Ding, Lianwen Jin
First submitted to arxiv on: 28 Feb 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper provides a comprehensive survey of Large Language Model (LLM) datasets, which are essential for the development and advancement of LLMs. The authors consolidate and categorize LLM dataset aspects from five perspectives: pre-training corpora, instruction fine-tuning datasets, preference datasets, evaluation datasets, and traditional NLP datasets. The survey highlights prevailing challenges and suggests potential avenues for future investigation. Additionally, a review of existing dataset resources is provided, including statistics from 444 datasets across 8 languages, spanning 32 domains, with information from 20 dimensions. The total data size surveyed exceeds 774.5 TB for pre-training corpora and 700M instances for other datasets. This paper aims to present the LLM text dataset landscape, serving as a comprehensive reference for researchers in this field. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper looks at the special collections of words and phrases that help large language models learn and improve. These collections are like a root system that helps the models grow strong. The authors want to understand what these collections contain and how they can be used better. They group the collections into five categories: training data, fine-tuning data, preference data, evaluation data, and traditional language processing data. By studying these collections, researchers can learn more about how to make large language models work better. |
Keywords
» Artificial intelligence » Fine tuning » Large language model » Nlp
