AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
How to Evaluate a Language Model?
I think open air is really great in carefully curating high quality data set. That something that they are very well perfect. But then on the other hand, there is also the question how you evaluate the model. And so in big science, we made the choice of trying to have data said that is representative of what people are actually reading on the internet. The evaluation was also a little bit more than forty working group in big siencs.