☆ Yσɠƚԋσʂ ☆@lemmy.ml to Machine Learning@lemmy.ml • edit-24 months agoBy using the same techniques Google used to solve Go (MTCS and backprop), Llama8B gets 96.7% on math benchmark GSM8K. That’s better than GPT-4, Claude and Gemini, with 200x fewer parameters!external-linkmessage-square0 fedilinkarrow-up17file-textcross-posted to: technology@lemmygrad.ml
arrow-up17external-linkBy using the same techniques Google used to solve Go (MTCS and backprop), Llama8B gets 96.7% on math benchmark GSM8K. That’s better than GPT-4, Claude and Gemini, with 200x fewer parameters!☆ Yσɠƚԋσʂ ☆@lemmy.ml to Machine Learning@lemmy.ml • edit-24 months agomessage-square0 Commentsfedilinkfile-textcross-posted to: technology@lemmygrad.ml