Update readme

.vscode/settings.json

@@ -5,6 +5,8 @@
     "downsample",
     "downsampled",
     "downsampler",
+    "downscaled",
+    "downscaler",
     "Eding",
     "getdata",
     "getpixel",

README.md

@@ -28,9 +28,7 @@ There is no preprocessor. Instead, supply a black and white checkerboard image a
 
 The script `gen_checker.py` can be used to generate checkerboard images of arbitrary sizes. (https://huggingface.co/thomaseding/pixelnet/blob/main/gen_checker.py) Example: `python gen_checker.py --upscale-dims 512x512 --dims 70x70 --output-file control.png` to generate a 70x70 checkerboard image upscaled to 512x512 pixels.
 
-![grid5x5](https://huggingface.co/thomaseding/pixelnet/resolve/main/example-control-images/5x5.png)
-
-![grid16x16](https://huggingface.co/thomaseding/pixelnet/resolve/main/example-control-images/16x16.png)
+The script `controlled_downscale.py` is a custom downscaler made specifically for this model. You provide both the generated image and the control image used to generate it. It will downscale according to the control grid. (https://huggingface.co/thomaseding/pixelnet/blob/main/controlled_downscale.py) Example: `python controlled_downscale.py --control diffusion_control.png --input diffusion_output.png --output-downscaled downscaled.png --output-quantized quantized.png --trim-cropped-edges true`
 
 ### FAQ:
 
@@ -40,7 +38,7 @@ A: From my experience SD has a hard time creating genuine pixel art (even with d
 
 Q: Should I use this model with a post-processor?
 
-A: Yes, I still recommend you do post-processing to clean up the image. This model is not perfect and will still have artifacts. Note that none of the sample output images are post-processed; they are raw outputs from the model. Consider sampling the image based on the location of the control grid checker faces. I will provide a custom script specialized for this in the near future.
+A: Yes, I still recommend you do post-processing to clean up the image. This model is not perfect and will still have artifacts. Note that none of the sample output images are post-processed; they are raw outputs from the model. Consider sampling the image based on the location of the control grid checker faces. The provided `controlled_downscale.py` script can do this for you. You can take the output of this script (presumably the `--output-downscaled` file) and then run it through a different post-processor (e.g. to refine the color palette).
 
 Q: Does the model support non-square grids?
 A: Kind of. I trained it with some non-perfect square grids (when pre-upscaled checkerboards are not a factor of the upscaled image size), so in that sense it should work fine. I also trained it with some checkerboard images with genuine non-square rectangular faces (e.g. double-wide pixels).