Which hardware should run this model? Most teams answer that question with habit instead of arithmetic. The model card says GPU. The cluster has a GPU quota. Th
