数据规模与模型规模 #8
-
用200G 的数据训练10B 的模型,这个规模上是否匹配呢?数据量是不是有点少了呢? |
Beta Was this translation helpful? Give feedback.
Answered by
jayzzhou-thu
May 26, 2022
Replies: 1 comment 5 replies
-
您好, |
Beta Was this translation helpful? Give feedback.
5 replies
Answer selected by
zh-zheng
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
您好,
175B的GPT3使用了大概300B Token的数据,目前我们清洗之后的200G数据大约是50B的Token数量,用来训练10B的模型在量级上应该是足够的。
此外200G目前也是初始阶段的数据量,后续有新的数据我们也会不断增加~
感谢您的提问!