Update app.py

app.py

@@ -45,7 +45,7 @@ latent_dim = 512
 
 # Load original generator
 original_generator = Generator(1024, latent_dim, 8, 2).to(device)
-ckpt = torch.load(os.path.join('models', ckpt), map_location=lambda storage, loc: storage)
+ckpt = torch.load(os.path.join('models', 'stylegan2-ffhq-config-f.pt'), map_location=lambda storage, loc: storage)
 original_generator.load_state_dict(ckpt["g_ema"], strict=False)
 mean_latent = original_generator.mean_latent(10000)
 
@@ -61,68 +61,49 @@ transform = transforms.Compose(
 )
 
 plt.rcParams['figure.dpi'] = 150
-
-filepath = f'test_input/{filename}'
-
-name = strip_path_extension(filepath)+'.pt'
-
-aligned_face = align_face(filepath)
-
 os.system("gdown https://drive.google.com/uc?id=1o6ijA3PkcewZvwJJ73dJ0fxhndn0nnh7")
 os.system("mv e4e_ffhq_encode.pt models/e4e_ffhq_encode.pt")
-my_w = e4e_projection(aligned_face, name, device).unsqueeze(0)
 
 
-plt.rcParams['figure.dpi'] = 150
-pretrained = 'jojo' #@param ['supergirl', 'arcane_jinx', 'arcane_caitlyn', 'jojo_yasuho', 'jojo', 'disney']
-#@markdown Preserve color tries to preserve color of original image by limiting family of allowable transformations. Otherwise, the stylized image will inherit the colors of the reference images, leading to heavier stylizations.
-preserve_color = False #@param{type:"boolean"}
-
-if preserve_color:
-    ckpt = f'{pretrained}_preserve_color.pt'
-else:
-    ckpt = f'{pretrained}.pt'
-
-downloader.download_file(ckpt)
-ckpt = torch.load(os.path.join('models', ckpt), map_location=lambda storage, loc: storage)
-generator.load_state_dict(ckpt["g"], strict=False)
-
-#@title Generate results
-n_sample =  1#@param {type:"number"}
-seed = 3000 #@param {type:"number"}
-
-torch.manual_seed(seed)
-with torch.no_grad():
-    generator.eval()
-    z = torch.randn(n_sample, latent_dim, device=device)
-
-    original_sample = original_generator([z], truncation=0.7, truncation_latent=mean_latent)
-    sample = generator([z], truncation=0.7, truncation_latent=mean_latent)
-
-    original_my_sample = original_generator(my_w, input_is_latent=True)
-    my_sample = generator(my_w, input_is_latent=True)
-
-# display reference images
-style_path = f'style_images_aligned/{pretrained}.png'
-style_image = transform(Image.open(style_path)).unsqueeze(0).to(device)
-face = transform(aligned_face).unsqueeze(0).to(device)
-
-my_output = torch.cat([style_image, face, my_sample], 0)
-display_image(utils.make_grid(my_output, normalize=True, range=(-1, 1)), title='My sample')
-
-output = torch.cat([original_sample, sample], 0)
-display_image(utils.make_grid(output, normalize=True, range=(-1, 1), nrow=n_sample), title='Random samples')
-
-def inference(img, ver):
-    if ver == 'version 2 (🔺 robustness,🔻 stylization)':
-        out = face2paint(model2, img)
-    else:
-        out = face2paint(model1, img)
-    return out
+os.system("gdown https://drive.google.com/uc?id=13cR2xjIBj8Ga5jMO7gtxzIJj2PDsBYK4")
+os.system("mv e4e_ffhq_encode.pt models/jojo.pt")
+
+os.system("gdown https://drive.google.com/uc?id=1ZRwYLRytCEKi__eT2Zxv1IlV6BGVQ_K2")
+os.system("mv e4e_ffhq_encode.pt models/jojo_preserve_color.pt")
+
+def inference(img):
+    filepath = img
+    
+    name = strip_path_extension(filepath)+'.pt'
+    
+    aligned_face = align_face(filepath)
+    
+    my_w = e4e_projection(aligned_face, name, device).unsqueeze(0)
+    
+    
+    plt.rcParams['figure.dpi'] = 150
+    pretrained = 'jojo' #@param ['supergirl', 'arcane_jinx', 'arcane_caitlyn', 'jojo_yasuho', 'jojo', 'disney']
+    #@markdown Preserve color tries to preserve color of original image by limiting family of allowable transformations. Otherwise, the stylized image will inherit the colors of the reference images, leading to heavier stylizations.
+    preserve_color = False #@param{type:"boolean"}
+    
+
+    ckpt = torch.load(os.path.join('models', 'jojo.pt'), map_location=lambda storage, loc: storage)
+    generator.load_state_dict(ckpt["g"], strict=False)
+    
+    with torch.no_grad():
+        generator.eval()
+    
+        original_my_sample = original_generator(my_w, input_is_latent=True)
+        my_sample = generator(my_w, input_is_latent=True)
+    
+    return 
+   
+    my_output = torch.cat([style_image, face, my_sample], 0)   
+
+    return my_output
   
 title = "AnimeGANv2"
 description = "Gradio Demo for AnimeGanv2 Face Portrait. To use it, simply upload your image, or click one of the examples to load them. Read more at the links below. Please use a cropped portrait picture for best results similar to the examples below."
 article = "<p style='text-align: center'><a href='https://github.com/bryandlee/animegan2-pytorch' target='_blank'>Github Repo Pytorch</a></p> <center><img src='https://visitor-badge.glitch.me/badge?page_id=akhaliq_animegan' alt='visitor badge'></center> <p style='text-align: center'>samples from repo: <img src='https://user-images.githubusercontent.com/26464535/129888683-98bb6283-7bb8-4d1a-a04a-e795f5858dcf.gif' alt='animation'/> <img src='https://user-images.githubusercontent.com/26464535/137619176-59620b59-4e20-4d98-9559-a424f86b7f24.jpg' alt='animation'/><img src='https://user-images.githubusercontent.com/26464535/127134790-93595da2-4f8b-4aca-a9d7-98699c5e6914.jpg' alt='animation'/></p>"
-examples=[['groot.jpeg','version 2 (🔺 robustness,🔻 stylization)'],['bill.png','version 1 (🔺 stylization, 🔻 robustness)'],['tony.png','version 1 (🔺 stylization, 🔻 robustness)'],['elon.png','version 2 (🔺 robustness,🔻 stylization)'],['IU.png','version 1 (🔺 stylization, 🔻 robustness)'],['billie.png','version 2 (🔺 robustness,🔻 stylization)'],['will.png','version 2 (🔺 robustness,🔻 stylization)'],['beyonce.png','version 1 (🔺 stylization, 🔻 robustness)'],['gongyoo.jpeg','version 1 (🔺 stylization, 🔻 robustness)']]
-gr.Interface(inference, [gr.inputs.Image(type="pil"),gr.inputs.Radio(['version 1 (🔺 stylization, 🔻 robustness)','version 2 (🔺 robustness,🔻 stylization)'], type="value", default='version 2 (🔺 robustness,🔻 stylization)', label='version')
-], gr.outputs.Image(type="pil"),title=title,description=description,article=article,enable_queue=True,examples=examples,allow_flagging=False).launch()
+
+gr.Interface(inference, [gr.inputs.Image(type="file")], gr.outputs.Image(type="pil"),title=title,description=description,article=article,enable_queue=True,allow_flagging=False).launch()