Choose all correct descriptions that describe the Vision Transformer $($ViT$).$

$1.$ One of the disadvantages of the Vision Transformer is its inability to handle large$-$scale datasets effectively.

$2.$ The Vision Transformer typically requires a larger amount of data for training compared to convolutional neural networks.

$3.$ The Vision Transformer model processes input images by dividing them into fixed$-$size patches.

$4.$ The Vision Transformer eliminates the need for convolutional layers entirely.