Evaluating Language Models
The Data Exchange with Ben Lorica
00:00
The Future of Helm: A Community Initiative
Helm is Open Source. Everything's Open Source. If you have a data set that you're interested in, maybe your unique scenarios that aren't being represented, we'd love to accept contributions. And if they can add it to the Helm benchmark, then we're basically running these data sets on all the models. I hope that this will be a community effort where we build up this and gradually fleshing out the evaluation space.
Transcript
Play full episode
