trial

app.py

@@ -1,20 +1,22 @@
 import os
 os.system("pip install gradio==2.4.6")
-import gradio as gr
-from PIL import Image
-import torch
+os.system("pip install gdown lpips")
+os.system("gdown --id 1HKmjg6iXsWr4aFPuU0gBXPGR83wqMzq7 -O align.dat")
+os.system("wget https://nvlabs-fi-cdn.nvidia.com/stylegan2-ada-pytorch/pretrained/ffhq.pkl")
+os.system("gdown https://github.com/ninja-build/ninja/releases/download/v1.10.2/ninja-linux.zip")
+os.system("unzip -d /usr/local/bin/")
+os.system("sudo update-alternatives --install /usr/bin/ninja ninja /usr/local/bin/ninja 1 --force")
+os.mkdir("embeddings/")
 
 
-def greet(num):
-    return num+69
+import gradio as gr
 
-iface = gr.Interface(fn=greet, inputs="number", outputs="number")
-iface.launch(share=True)
 
 
 def inference(img):
-    out = face2paint(model1, img)
-    return out
+    img.save("images/file.png")
+    os.system("python tune.py")
+    return 
   
 title = "Pivotal Tuning for Latent Based Real Image Editing"
 description = "Gradio Demo for Pivotal Tuning Inversion. 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."

configs/paths_config.py

@@ -1,9 +1,9 @@
 ## Pretrained models paths
-e4e = './pretrained_models/e4e_ffhq_encode.pt'
-stylegan2_ada_ffhq = '/home/sayantan/PTI/pretrained_models/ffhq.pkl'
+e4e = 'e4e_ffhq_encode.pt'
+stylegan2_ada_ffhq = 'ffhq.pkl'
 style_clip_pretrained_mappers = ''
-ir_se50 = './pretrained_models/model_ir_se50.pth'
-dlib = './pretrained_models/align.dat'
+ir_se50 = 'model_ir_se50.pth'
+dlib = 'align.dat'
 
 ## Dirs for output files
 checkpoints_dir = './checkpoints'

requirements.txt

@@ -5,5 +5,4 @@ gdown
 numpy
 scipy
 cmake
-onnxruntime-gpu
-opencv-python-headless
+onnxruntime-gpu

tune.py

@@ -13,24 +13,19 @@ from scripts.latent_editor_wrapper import LatentEditorWrapper
 image_dir_name = 'images'
 use_multi_id_training = False
 global_config.device = 'cuda'
-paths_config.e4e = '/home/sayantan/PTI/pretrained_models/e4e_ffhq_encode.pt'
+paths_config.e4e = 'e4e_ffhq_encode.pt'
 paths_config.input_data_id = image_dir_name
 paths_config.input_data_path = f'{image_dir_name}'
-paths_config.stylegan2_ada_ffhq = '/home/sayantan/PTI/pretrained_models/ffhq.pkl'
-paths_config.checkpoints_dir = '/home/sayantan/PTI/'
-paths_config.style_clip_pretrained_mappers = '/home/sayantan/PTI/pretrained_models'
+paths_config.stylegan2_ada_ffhq = 'ffhq.pkl'
+paths_config.checkpoints_dir = ''
+paths_config.style_clip_pretrained_mappers = ''
 hyperparameters.use_locality_regularization = False
 hyperparameters.lpips_type = 'squeeze'
 
 from scripts.run_pti import run_PTI
 
-@click.command()
-@click.pass_context
-@click.option('--rname', prompt='wandb RUN NAME', help='The name to give for the wandb run')
-
-def tune(ctx: click.Context,rname):
-    runn = wandb.init(project='PTI', entity='masc', name = rname)
-    model_id = run_PTI(run_name='',use_wandb=True, use_multi_id_training=False)
+def tune():
+    model_id = run_PTI(run_name='',use_wandb=False, use_multi_id_training=False)
 
 #----------------------------------------------------------------------------
 if __name__ == '__main__':