Update app.py

app.py

@@ -6,6 +6,7 @@ import gradio as gr
 import numpy as np
 import PIL.Image
 import torch
+import torchvision.transforms.functional as TF
 from diffusers import DDPMScheduler, StableDiffusionXLAdapterPipeline, T2IAdapter
 
 DESCRIPTION = "# T2I-Adapter-SDXL Sketch"
@@ -16,12 +17,7 @@ if not torch.cuda.is_available():
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 if torch.cuda.is_available():
     model_id = "stabilityai/stable-diffusion-xl-base-1.0"
-    adapter = T2IAdapter.from_pretrained(
-        "Adapter/t2iadapter",
-        subfolder="sketch_sdxl_1.0",
-        torch_dtype=torch.float16,
-        adapter_type="full_adapter_xl",
-    )
+    adapter = T2IAdapter.from_pretrained("TencentARC/t2i-adapter-sketch-sdxl-1.0", torch_dtype=torch.float16, variant="fp16")
     scheduler = DDPMScheduler.from_pretrained(model_id, subfolder="scheduler")
     pipe = StableDiffusionXLAdapterPipeline.from_pretrained(
         model_id,
@@ -48,12 +44,14 @@ def run(
     image: PIL.Image.Image,
     prompt: str,
     negative_prompt: str,
-    num_steps=50,
+    num_steps=25,
     guidance_scale=7.5,
+    adapter_conditioning_scale=0.8
     seed=0,
 ) -> PIL.Image.Image:
-    # Convert the input image, which is a boolean image, to a grayscale image whose value is 0 or 255.
-    image = image.convert("L")
+    image = image.convert("RGB").resize((1024, 1024))
+    image = TF.to_tensor(image) > 0.5
+    image = TF.to_pil_image(image.to(torch.float32))
 
     generator = torch.Generator(device=device).manual_seed(seed)
     out = pipe(
@@ -63,6 +61,7 @@ def run(
         num_inference_steps=num_steps,
         generator=generator,
         guidance_scale=guidance_scale,
+        adapter_conditioning_scale=adapter_conditioning_scale,
     ).images[0]
     return out