starts with conv layer then has `num_blocks` `ResLayer`

Args: depth (int): depth of darknet used in model, usually use [21, 53] for this param. in_channels (int): number of

We resize both tensors to [A,B,2] without new malloc: [A,2] -> [A,1,2] -> [A,B,2] [B,2] -> [1,B,2] -> [A,B,2] Then we compute the area of

check if model is in parallel mode.

Args: model (nn.Module): model to apply EMA. decay (float): ema decay reate. updates (int): counter of EM

Args: dataloader (Dataloader): evaluate dataloader. img_size (int): image size after preprocess. images are resized

Args: in_channels (int): input channels. out_channels (int): output channels. n (int): number of Bottlene

Args: act (str): activation type of conv. Defalut value: "silu". depthwise (bool): whether apply depthwise conv in co

Checks if the center of box_a is in box_b. Return: (tensor) int tensor of zeros and ones.

output: (N, 5+class) shape