Post AZvFr4wQA6eRBSAMzI by daridrea@graphics.social
 (DIR) More posts by daridrea@graphics.social
 (DIR) Post #AZvFr4wQA6eRBSAMzI by daridrea@graphics.social
       2023-09-19T10:42:25Z
       
       0 likes, 1 repeats
       
       Würstchen: "...achieves notable speed improvements at inference time, thereby rendering real-time applications more viable. One of the key advantages of our method lies in its modest training requirements of only 9,200 GPU hours, slashing the usual costs significantly without compromising the end performance." https://arxiv.org/abs/2306.00637