Abdelrahman Shaker

Mobile-VideoGPT is an efficient multimodal framework designed to operate with fewer than a billion parameters and real-time throughput.

This collection includes the pre-training, sft, and post-training data of Mobile-O