I have one channel images and want to add them to the tensorboard logger by using writer.add_image, as follows:

 generated_images_to_show = torch.cat(gtimage_cpu, predimage_cpu), dim=0)
 gen_image_grid = make_grid(generated_images_to_show .unsqueeze(1), nrow=inputs.size(0), padding=6, normalize=False)
 writer.add_image('Validation Gen Comparison', gen_image_grid, epoch)

My generated images have one channel and I need to plot them with cmap= 'hsv' but no cmap feature is available in add_image. In this case, what should I do? I am interested in viewing the results in tensorboard only.

One solution would be converting the images to RGB-channels and then use cmap=hsv and log the converted images to the tensorboard, is there a different (perhaps more efficient) solution?