Function get_ancestral_step

k_diffusion/sampling.py:51–58

Calculates the noise level (sigma_down) to step down to and the amount of noise to add (sigma_up) when doing an ancestral sampling step.

(sigma_from, sigma_to, eta=1.)

def get_ancestral_step(sigma_from, sigma_to, eta=1.):
    """Calculates the noise level (sigma_down) to step down to and the amount
    of noise to add (sigma_up) when doing an ancestral sampling step."""
    if not eta:
        return sigma_to, 0.
    sigma_up = min(sigma_to, eta * (sigma_to ** 2 * (sigma_from ** 2 - sigma_to ** 2) / sigma_from ** 2) ** 0.5)
    sigma_down = (sigma_to ** 2 - sigma_up ** 2) ** 0.5
    return sigma_down, sigma_up


def default_noise_sampler(x):