Robin V2 Launches: Achieves Unparalleled Performance on OpenLLM!
A stunning arrival! The fully upgraded Robin Series V2 language model is ready and eagerly awaiting your exploration.
This is not just a model upgrade, but the crystallization of wisdom from our research and development team. In the new version, Robin Series V2 has performed excellently among various open-source models, defeating well-known models such as Falcon, LLaMA, StableLM, RedPajama, MPT.
Specifically, we have carried out in-depth fine-tuning based on the entire LLaMA series, including 7b, 13b, 33b, 65b, all of which have achieved pleasing results. Robin-7b scored 51.7 in the OpenLLM standard test, and Robin-13b even reached as high as 59.1, ranking sixth, surpassing many 33b models. The achievements of Robin-33b and Robin-65b are even more surprising, with scores of 64.1 and 65.2 respectively, firmly securing the top positions.
It’s worth mentioning that we have also proposed a brand new evaluation scheme, utilizing negative log-likelihood to calculate the model’s performance capability in different domains, and the Robin model has significant advantages in multiple fields. For detailed evaluation results of the model, you can check in the LMFlow-benchmark.
In addition, to make it easier for you to develop and use the Robin Series V2, we provide a full set of supporting tools, LMFlow. Whether it’s pre-training, fine-tuning, RLHF, or evaluation, all can be done in one-stop in LMFlow.
We welcome you to join the ranks of Robin Series V2 users, to experience the powerful performance and rich features of this super model. If you have any questions or suggestions during use, you are always welcome to contact us. We have dedicated services set up in slack, discord, and WeChat community groups to answer your questions at any time.
Join us quickly and feel the surprise that Robin Series V2 brings to you!
