Confuse in the 'Sample Fuction'

[ankan8145]

def sample(self, model, n):
        logging.info(f"Sampling {n} new images....")
        model.eval()
        with torch.no_grad():
            x = torch.randn((n, 3, self.img_size, self.img_size)).to(self.device)
            for i in tqdm(reversed(range(1, self.noise_steps)), position=0):
                t = (torch.ones(n) * i).long().to(self.device)
                predicted_noise = model(x, t)
                alpha = self.alpha[t][:, None, None, None]
                alpha_hat = self.alpha_hat[t][:, None, None, None]
                beta = self.beta[t][:, None, None, None]
                if i > 1:
                    noise = torch.randn_like(x)
                else:
                    noise = torch.zeros_like(x)
                x = 1 / torch.sqrt(alpha) * (x - ((1 - alpha) / (torch.sqrt(1 - alpha_hat))) * predicted_noise) + torch.sqrt(beta) * noise
        model.train()
        x = (x.clamp(-1, 1) + 1) / 2
        x = (x * 255).type(torch.uint8)
        return x 

Can anyone explain this fuction . In the line 'x = torch.randn((n, 3, self.img_size, self.img_size)).to(self.device)', you create a random image then from that image you predict the noise ( i.e. predicted_noise = model(x, t) ). Are you tring to create an image from a random tensor ??

[randomaccess2023]

@ankan8145 x = torch.randn((n, 3, self.img_size, self.img_size)).to(self.device) is a random noise with n=12, channels=3 and self.img_size=64 (probably). Random sampling (after training the DDPM) is carried out by starting from pure noise and reversing the time steps (i.e., starting from 1000 and finishing at 1).

The outcome is realistic-looking images obtained from pure noise only.

[CourtPoet]

Do you need a numerical range of randomly generated noise according to the training set requirements?

[randomaccess2023]

@CourtPoet The randomly generated noise needs to be a Gaussian distribution. In PyTorch, this function can be used ---> torch.randn.