Differentiable rounding function

AFAIK, it's currently only possible to get the accent color programmatically. Other colors need apparently be to be hardcoded and pra

Use it only if producer we're waiting for can produce an unknown number of items before it stops and could therefore potentially block on put

Evaluate the model performance of the whole dataset after processing all batches. Different like :class:`~mmengine.evaluator.BaseMetric`,

Evaluate generative metric. In this function we only collect :attr:`fake_results` because generative metrics do not need real images.

Return the experiment object associated with this visualization backend.

Number of fake images need for current device.

Args: x (Tensor): Input image, bchw, rgb, [0, 1] quality(float): Quality factor for jpeg compression scheme.

Args: image(tensor): batch x 3 x height x width Returns: dict(tensor): Compressed tensor with batch x h*w/64

Args: image(Tensor): batch x 3 x height x width Returns: Tensor: batch x height x width x 3

Args: image(tensor): batch x height x width x 3 Returns: y(tensor): batch x height x width cb(ten

Args: image(tensor): batch x height x width Returns: Tensor: batch x h*w/64 x h x w