We will try to beat the Efficientnet-B4 as B7 is very resource-intensive to train. I trained the models on Colab Pro, which runs 16 GB ...
確定! 回上一頁