Update app.py

app.py

@@ -1,7 +1,46 @@
 import gradio as gr
+import torch, torchvision
+import torch.nn.functional as F
+import numpy as np
+from PIL import Image, ImageColor
+from diffusers import DDPMPipeline
+from diffusers import DDIMScheduler
 
-def greet(name):
-    return "Hello " + name + "!!"
+device = 'mps' if torch.backends.mps.is_available() else 'cuda' if torch.cuda.is_available() else 'cpu'
 
-iface = gr.Interface(fn=greet, inputs="text", outputs="text")
-iface.launch(share=True)
+# Load the pretrained pipeline
+pipeline_name = 'WiNE-iNEFF/Minecraft-Skin-Diffusion'
+image_pipe = DDPMPipeline.from_pretrained(pipeline_name).to(device)
+
+# Set up the scheduler
+scheduler = DDIMScheduler.from_pretrained(pipeline_name)
+scheduler.set_timesteps(num_inference_steps=40)
+
+# And the core function to generate an image given the relevant inputs
+def generate():
+    x = torch.randn(8, 4, 64, 64).to(device)
+    # Minimal sampling loop
+    for i, t in tqdm(enumerate(scheduler.timesteps)):
+        model_input = scheduler.scale_model_input(x, t)
+        with torch.no_grad():
+            noise_pred = image_pipe.unet(model_input, t)["sample"]
+        x = scheduler.step(noise_pred, t, x).prev_sample
+    # View the results
+    grid = torchvision.utils.make_grid(x, nrow=4)
+    im = grid.permute(1, 2, 0).cpu().clip(-1, 1) * 0.5 + 0.5
+    im.convert("RGBA").save("test.png")
+    return im
+
+# See the gradio docs for the types of inputs and outputs available
+outputs = gr.Image(label="result")
+
+# Setting up a minimal interface to our function:
+demo = gr.Interface(
+    fn=generate,
+    inputs=None,
+    outputs=outputs,
+)
+
+# And launching
+if __name__ == "__main__":
+    demo.launch(enable_queue=True)