INNER CODE UNIT · Python

batch_id_

KlingAIResearch/ReCamMaster · diffsynth/extensions/ESRGAN/__init__.py:122

            batch_id_ = min(batch_id + batch_size, input_tensor.shape[0])
            batch_input_tensor = input_tensor[batch_id: batch_id_]
            batch_input_tensor = batch_input_tensor.to(
                device=self.model.conv_first.weight.device,
                dtype=self.model.conv_first.weight.dtype)
            batch_output_tensor = self.model(batch_input_tensor)
            output_tensor.append(batch_output_tensor.cpu())
        
        # Output
        output_tensor = torch.concat(output_tensor, dim=0)

        # To images
        output_images = self.decode_images(output_tensor)
        if is_single_image:
            output_images = output_images[0]
        return output_images

View source record →