build: specify proper Pillow minimum version (fixes #325)

README.md

@@ -485,8 +485,8 @@ A: The AI models are cached in `~/.cache/` (or `HUGGINGFACE_HUB_CACHE`). To dele
 
 **13.0.0**
 - 🎉 feature: multi-controlnet support. pass in multiple `--control-mode`, `--control-image`, and `--control-image-raw` arguments.
-- 🎉 feature: "better" memory management. If GPU is full, least-recently-used model is moved to RAM.
 - 🎉 feature: add colorization controlnet. improve `aimg colorize` command
+- 🧪 feature: "better" memory management. If GPU is full, least-recently-used model is moved to RAM. I'm not confident this works well.
 - 🧪 alpha feature: `aimg run-api-server` command. Runs a http webserver (not finished). After running, visit http://127.0.0.1:8000/docs for api.
 - feature: [disabled] inpainting controlnet can be used instead of finetuned inpainting model
   - The inpainting controlnet doesn't work as well as the finetuned model
@@ -495,6 +495,7 @@ A: The AI models are cached in `~/.cache/` (or `HUGGINGFACE_HUB_CACHE`). To dele
 - fix: hide the "triton" error messages
 - fix: package will not try to install xformers on `aarch64` machines. While this will allow the dockerfile to build on 
 MacOS M1, [torch will not be able to use the M1 when generating images.](https://github.com/pytorch/pytorch/issues/81224#issuecomment-1499741152)
+- build: specify proper Pillow minimum version (fixes #325) 
 
 **12.0.3**
 - fix: exclude broken versions of timm as dependencies

setup.py

@@ -83,7 +83,7 @@ setup(
         "tqdm",
         "diffusers",
         "imageio>=2.9.0",
-        "Pillow>=8.0.0",
+        "Pillow>=9.1.0",
         "psutil",
         # 2.0.0 need to fix `ImportError: cannot import name 'rank_zero_only' from 'pytorch_lightning.utilities.distributed' `
         "pytorch-lightning>=1.4.2,<2.0.0",