M machinelearning.apple.com FastVLM: Efficient Vision Encoding for Vision Language Models 93 points 2bit 1 year ago 6 comments
