Trained on a much larger dataset (73 images) with better captioning, as well as fixing spelling errors! Much less overfitting as well.
Trained on a much larger dataset (73 images) with better captioning, as well as fixing spelling errors! Much less overfitting as well.