From ed6aff8caa3f3bf8438732922a9c9d07f21fd86a Mon Sep 17 00:00:00 2001 From: Lj Miranda Date: Sun, 6 Aug 2023 17:19:21 +0800 Subject: [PATCH] Fix spelling and grammar --- notebook/_posts/2023-08-04-llm-tagalog.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/notebook/_posts/2023-08-04-llm-tagalog.md b/notebook/_posts/2023-08-04-llm-tagalog.md index cb9a407b..ff94b59d 100644 --- a/notebook/_posts/2023-08-04-llm-tagalog.md +++ b/notebook/_posts/2023-08-04-llm-tagalog.md @@ -158,7 +158,7 @@ For example, the [Pile](https://pile.eleuther.ai/) dataset is English-only while Unfortunately, Tagalog is underrepresented even in multilingual LMs. For example, the XGLM language model ([Lin et al., 2022](#lin2022xglm)) was only trained on 2.3k tokens of Tagalog data whereas the Bloom language model ([Scao et al., 2022](#scao2022bloom)) doesn't contain any Tagalog text at all. There's still a long way to go. - Currently, there are efforts such as [Cohere's Aya project](https://txt.cohere.com/aya-multilingual/) that aims to close that multilingual gap. + Currently, there are efforts such as [Cohere's Aya project](https://txt.cohere.com/aya-multilingual/) that aim to close that multilingual gap. ### Thoughts on how we should use LLMs