N nexa.ai Omnivision-968M: Vision Language Model with 9x Tokens Reduction for Edge Devices 69 points BUFU 2 years ago 12 comments
