update README.md: more justifications for the bloat

README.md

@@ -9,7 +9,7 @@ tags:
 - stable-diffusion-xl
 ---
 
-This is the single file [`furry-xl-4.0.safetensors`](https://huggingface.co/SeaArtLab/SeaArt-Furry-XL-1.0) converted to diffusers format. A wolf asked for it. If you're not a wolf you might still want this because the original version is fp32 while this one is fp16.
+This is the single file [`furry-xl-4.0.safetensors`](https://huggingface.co/SeaArtLab/SeaArt-Furry-XL-1.0) [converted](https://github.com/Linaqruf/sdxl-model-converter/) to diffusers format. A wolf asked for it. If you're not a wolf you might still want this because the original version is fp32 while this one is fp16. SeaArt's diffusers distribution is also [missing the `unet/config.json`](https://huggingface.co/SeaArtLab/SeaArt-Furry-XL-1.0/discussions/2).
 
 # SeaArt Furry XL 1.0
 ![image/png](https://cdn-uploads.huggingface.co/production/uploads/66139f0f09f16e734729c269/gDx2bPiQzvzxUwfPdB_G8.jpeg)
@@ -33,7 +33,7 @@ import torch
 from diffusers import DiffusionPipeline
 
 pipe = DiffusionPipeline.from_pretrained(
-    "Gaeros/SeaArt-Furry-XL-1.0",
+    "Gaeros/SeaArt-Furry-XL-1.0-fp16-diffusers",
     torch_dtype=torch.float16,
 )
 pipe.to('cuda')