Upload 3 files

app.py

@@ -0,0 +1,38 @@
+import gradio as gr
+import torch
+import numpy as np
+import os
+from model import gen_model
+import torchvision.transforms as T
+
+# Model
+gen,transform_gen=gen_model()
+# print(gen)
+to_img=T.ToPILImage()
+# examples=["examples/input_0.png","examples/input_9.png"]
+example_list = [["examples/" + example] for example in os.listdir("examples")]
+
+# Predict Function
+def predict(img):
+  # Apply Transformations
+  img=np.array(img)
+  img=transform_gen(image=img)
+  img=img['image'].unsqueeze(0)
+  # Predict
+  gen.eval()
+  with torch.inference_mode():
+    y_gen=gen(img)
+  y_gen=y_gen[0]
+  y_gen=to_img(y_gen)
+  return y_gen
+
+# Gradio App
+title="Pix2Pix GAN"
+description="This is a Sattelite Image to Map converter"
+
+demo=gr.Interface(fn=predict,
+                  inputs=gr.Image(type='pil'),
+                  outputs=gr.Image(type='pil'),
+                  title=title ,
+                  examples=example_list,
+                  description=description)

model.py

@@ -0,0 +1,17 @@
+import torch
+# from torchvision import transforms
+
+
+def gen_model():
+    # transform = transforms.Compose([
+    # transforms.Resize((256, 256)),
+    # transforms.RandomCrop((224, 224)),
+    # transforms.RandomHorizontalFlip(),
+    # transforms.ToTensor(),
+    # transforms.Normalize(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5])])
+    device="cpu"
+    with torch.no_grad():
+        model = torch.load('gen.pth.tar', map_location='cpu')
+    model = model['state_dict']
+    model=model['state_dict']
+    return model,transform

requirements.txt

@@ -0,0 +1,4 @@
+torch==1.12.0
+torchvision==0.13.0
+gradio==3.14.0
+numpy==1.24.2