Antoine Aarts
Posted this news 1 year ago
open-source LLMs against a diverse set of benchmarks
ChatGPT launched one year ago today. It has ignited an arms race across both proprietary and open source model development, with the latter claiming parity or better on certain tasks.To test these claims, our Salesforce Research team (Hailin Chen Caiming Xiong et al.) evaluated the most popular and highly rated open-source LLMs against a diverse set of benchmarks... See more
