Models from the paper "No Labels, No Problem: Training Visual Reasoners with Multimodal Verifiers"