create an efficientnet model according to name. Args: model_name (str): Name for efficientnet. in_channels (int): Inp

create an efficientnet model according to name. Args: model_name (str): Name of PVT model, e.g. pvt_v2_b0 weights_pat

Gaussian Blurring, a Gaussian kernel is used. @param ksize: the kernel size of gaussian kernel, here the width and the height of

Generate ngrams for particular 'n' from a list of tokens Parameters ---------- tokens : List[str] List of tokens for which the ngr

Returns: The value of the key, if the key not exist, it will return a default value.

Args: grid: (B, *, H, W, d) Returns: (B, *, H, W, d)

Annotations of task, consists of all datasets' annotations

Return the corresponding `attr` attribute of the whole dataset as a list. This method provides an easy way for access datasets' attrs

Get batch size of the current ``SampleList``. There must be a tensor field present in the ``SampleList`` currently. Returns:

Get a new ``SampleList`` generated from the current ``SampleList`` but contains only the attributes passed in `fields` argument Args:

override in your dataset first