this post was submitted on 16 Jun 2024

8 points (78.6% liked)

Machine Learning

1742 readers

1 users here now

founded 4 years ago

MODERATORS

osipov@lemmy.ml

8

By 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! (arxiv.org)

submitted 3 months ago* (last edited 3 months ago) by yogthos@lemmy.ml to c/machinelearning@lemmy.ml

2 comments fedilink hide all child comments

top 2 comments

sorted by: hot top controversial new old

[–] shasta@lemm.ee 2 points 3 months ago (1 children)

Fewer

[–] yogthos@lemmy.ml 2 points 3 months ago

Fine