Omnivision-968M: Vision Language Model with 9x Tokens Reduction for Edge Devices(nexa.ai)55 points by BUFU 1 year ago | 8 comments