Args: img (PIL.Image or numpy.ndarray): List of images to be cropped in format (h, w, c) in numpy.ndarray Returns:

Create a multi-headed attention layer. :param num_heads: the number of heads :param size: model size (must be divisible by nu

Initializes position-wise feed-forward layer. :param input_size: dimensionality of the input. :param ff_size: dimensionality

Initializes the Transformer. :param hidden_size: hidden size and size of embeddings :param ff_size: position-wise feed-forwar

inline function for validation check

Computes multi-headed attention. :param k: keys [B, M, D] with M being the sentence length. :param v: values [B, M, D]

Args: - src_feats: (max_src_len, batch_size, D) - src_lens: (batch_size) Returns: - outputs: (max

Return the output size :return: