Returns the number of symbols in the dictionary

convert patch embedding weight from manual patchify + linear proj to conv

convert patch embedding weight from manual patchify + linear proj to conv

Initialize the weights

Workaround for ModelEma._load_checkpoint to accept an already-loaded object

Workaround for ModelEma._load_checkpoint to accept an already-loaded object

Load weights from a checkpoint. Args: checkpoint (Any): checkpoint contains the weights. Returns: `

Modified focal loss. Exactly the same as CornerNet. Runs faster and costs a little bit more memory Arguments: pred (batch x c x h x w

Prunes heads of the model. heads_to_prune: dict of {layer_num: list of heads to prune in this layer} See base class PreTrainedModel

Getting sliced (dim=0) tensor by mask. Supporting tensor and list/dict of tensors.

Add a possible candidate to the best neighbors * @param dist distance for that neighbor * @param index index of that neighbor */

Add a possible candidate to the best neighbors * @param dist distance for that neighbor * @param index index of that neighbor */

Add a possible candidate to the best neighbors * @param dist distance for that neighbor * @param index index of that neighbor */

Add a possible candidate to the best neighbors * @param dist distance for that neighbor * @param index index of that neighbor */

FabMap2 builds the inverted index and requires an additional training/test add function

FabMap2 builds the inverted index and requires an additional training/test add function