Unverified50% confidenceFactExact time
Books3 is a dataset comprising approximately 196,000 English-language books with a total size of about 37 GB, and has been widely used for training large language models.
1
Sources
50%
Confidence
Long-term
Relevance
9/6/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedA 10-page academic paper may consume 8,000–15,000 tokens when processed by a large language model.63% similarUnverified最新的大语言模型支持12.8万到100万token的上下文，但处理数十万字长篇小说时仍会出现信息遗忘和前后矛盾62% similarVerifiedThe internet's supply of high-quality text data for training large language models has been largely consumed.61% similarVerified大语言模型通过在数千亿乃至数万亿个词元（Token）的文本数据上进行自监督预训练59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/866137API
curl https://kongchang.com/api/v1/knowledge/claims/866137MCP
get_claim(id=866137)