News for hackers. Stuff that matters.

A comment on Cerebras-GPT vs. LLaMA AI Model Performance Comparison. Parent comment

Re:Cerebras-GPT vs. LLaMA AI Model Performance Comparison by MuffinFlavored on Wednesday March 29, 2023 @07:38PM (#35365518)
as an LLM how do we train it to know when to call a “math” plugin/API? (feed its input, go get output)
[ Parent | on HN ]
Re:Cerebras-GPT vs. LLaMA AI Model Performance Comparison by sebzim4500 on Wednesday March 29, 2023 @07:41PM (#35365563)
Either finetune with a few thousand example, or just tell it in the prompt to call the API when it needs to do maths (in context learning).
[ Parent | on HN ]
Re:Cerebras-GPT vs. LLaMA AI Model Performance Comparison by discordance on Wednesday March 29, 2023 @07:53PM (#35365702)
In the pipeline, first run a classification model to determine what class of question is being asked.

Then engage the most appropriate plugins.

[ Parent | on HN ]