VinciCoder: Unifying Multimodal Code Generation via Coarse-to-fine Visual Reinforcement Learning

Re-assign community