Without changing the architecture in a significant manner/using more training data, there does come a point where adding more parameters will result in no gain (and sometimes even worse results). There's also a…
Probably not much at all. They used a pretty straightforward and easy to run model, EfficientNetV2. Something that doesn't give out the best results, but will offer very quick inference and train time.
Without changing the architecture in a significant manner/using more training data, there does come a point where adding more parameters will result in no gain (and sometimes even worse results). There's also a…
Probably not much at all. They used a pretty straightforward and easy to run model, EfficientNetV2. Something that doesn't give out the best results, but will offer very quick inference and train time.