Holistic Evaluation of Language Models (HELM), a framework to increase the transparency of language models (https://arxiv.org/abs/2211.09110). This framework is also used to evaluate text-to-image models in Holistic Evaluation of Text-to-Image Models (HEIM) (https://arxiv.org/abs/2311.04287).
By convention, we save space by not putting the organization name in the model name. The command models did not follow this convention before, so we fix it to follow the convention.
By convention, we save space by not putting the organization name in the model name. The command models did not follow this convention before, so we fix it to follow the convention.