Unverified50% confidenceBenchmarkExact time
Chinchilla, trained with 70 billion parameters and 1.4 trillion tokens of training data, surpassed Gopher—a model with four times its parameter count—on multiple benchmarks.
1
Sources
50%
Confidence
Long-term
Relevance
9/5/2026
First Seen
Sources
Related Entities
Related Claims
Unverified2022年Hoffman等人的Chinchilla论文重新校准了算力最优的参数-数据比例，指出此前许多大模型训练不充分67% similarUnverified根据DeepMind在2022年提出的Chinchilla缩放定律，一个700亿参数的模型需要约1.4万亿token才能达到计算最优65% similarUnverifiedMeta's LLaMA paper disclosed that approximately 67% of its training data came from Common Crawl, of which only about 15% of the original data was retained after multiple rounds of filtering.64% similarUnverifiedChinchilla研究发现在固定计算预算下参数量和训练token数应按大致1:20比例匹配64% similarUnverified对LLaMA-7B所有注意力层应用r=8的LoRA，可训练参数量从70亿降至约400万，压缩比超过1750倍64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/859318API
curl https://kongchang.com/api/v1/knowledge/claims/859318MCP
get_claim(id=859318)