Image Captioning with ViT and GPT 2 Base

[sd003]

so if I try out more advanced models then the performance may improve because as far as my knowledge goes there is not any fundamental problem in the training process nor with the model itself …