Training data details - sources and license

#11
by nramrakhiyani - opened

Will it be possible for the model providers to throw some light on details about the data used to train the model? I request details such as dataset sources and licenses. The primary reason being that even though the model is "mit" licensed (and hence commercial friendly), unfriendly dataset licenses may also be a hurdle for commercial deployments.

Sign up or log in to comment