|
import click |
|
import os |
|
import sys |
|
import pickle |
|
import numpy as np |
|
from PIL import Image |
|
import torch |
|
from configs import paths_config, hyperparameters, global_config |
|
from IPython.display import display |
|
import matplotlib.pyplot as plt |
|
from scripts.latent_editor_wrapper import LatentEditorWrapper |
|
|
|
image_dir_name = 'images' |
|
use_multi_id_training = False |
|
global_config.device = 'cuda' |
|
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 = 'ffhq.pkl' |
|
paths_config.checkpoints_dir = 'checkpoints' |
|
paths_config.style_clip_pretrained_mappers = '' |
|
hyperparameters.use_locality_regularization = False |
|
hyperparameters.lpips_type = 'squeeze' |
|
|
|
from scripts.run_pti import run_PTI |
|
|
|
def load_generator(model_id): |
|
with open(f'{paths_config.checkpoints_dir}/model_{model_id}_file.pt', 'rb') as f_new: |
|
new_G = torch.load(f_new).cuda() |
|
return new_G |
|
|
|
def tensor_to_pil(img): |
|
img = (img.permute(0, 2, 3, 1) * 127.5 + 128).clamp(0, 255).to(torch.uint8).detach().cpu().numpy()[0] |
|
plt.axis('off') |
|
resized_image = Image.fromarray(img,mode='RGB').resize((256,256)) |
|
return resized_image |
|
|
|
def tune(): |
|
model_id = run_PTI(run_name='',use_wandb=False, use_multi_id_training=False) |
|
w_path_dir = f'{paths_config.embedding_base_dir}/{paths_config.input_data_id}' |
|
embedding_dir = f'{w_path_dir}/{paths_config.pti_results_keyword}/{image_name}' |
|
w_pivot = torch.load(f'{embedding_dir}/0.pt') |
|
new_G = load_generator(model_id) |
|
new_image = new_G.synthesis(w_pivot, noise_mode='const', force_fp32 = True) |
|
tensor_to_pil(new_image).save("output/out.png") |
|
|
|
|
|
if __name__ == '__main__': |
|
tune() |
|
|
|
|
|
|