### Copyright (C) 2020 Roy Or-El. All rights reserved. ### Licensed under the CC BY-NC-SA 4.0 license (https://creativecommons.org/licenses/by-nc-sa/4.0/legalcode). import argparse import os from util import util import torch class BaseOptions(): def __init__(self): self.parser = argparse.ArgumentParser() self.initialized = False def initialize(self): # experiment specifics self.parser.add_argument('--name', type=str, default='males_model', help='name of the experiment. It decides where to store samples and agingModels') self.parser.add_argument('--gpu_ids', type=str, default='0', help='gpu ids: e.g. 0 0,1,2, 0,2. use -1 for CPU') self.parser.add_argument('--checkpoints_dir', type=str, default='./checkpoints', help='agingModels are saved here') # input/output sizes self.parser.add_argument('--batchSize', type=int, default=1, help='input batch size') self.parser.add_argument('--loadSize', type=int, default=256, help='scale images to this size') self.parser.add_argument('--fineSize', type=int, default=256, help='then crop to this size') self.parser.add_argument('--input_nc', type=int, default=3, help='# of input image channels') self.parser.add_argument('--output_nc', type=int, default=3, help='# of output image channels') # for setting inputs self.parser.add_argument('--dataroot', type=str, default='./static/uploads/') self.parser.add_argument('--sort_classes', type=bool, default=True, help='a flag that indicates whether to sort the classes') self.parser.add_argument('--sort_order', type=str, default='0-2,3-6,7-9,15-19,30-39,50-69', help='a specific order to sort the classes, must contain all classes, only works when sort_classes is true') self.parser.add_argument('--resize_or_crop', type=str, default='resize_and_crop', help='scaling and cropping of images at load time [resize_and_crop|crop|scale_width|scale_width_and_crop]') self.parser.add_argument('--serial_batches', action='store_true', help='if true, takes images in order to make batches, otherwise takes them randomly') self.parser.add_argument('--no_flip', action='store_true', help='if specified, do not flip the images for data argumentation') self.parser.add_argument('--nThreads', default=4, type=int, help='# threads for loading data') self.parser.add_argument('--max_dataset_size', type=int, default=float("inf"), help='Maximum number of samples allowed per dataset. If the dataset directory contains more than max_dataset_size, only a subset is loaded.') self.parser.add_argument('--display_single_pane_ncols', type=int, default=6, help='if positive, display all images in a single visdom web panel with certain number of images per row.') # for displays self.parser.add_argument('--display_winsize', type=int, default=256, help='display window size') self.parser.add_argument('--display_port', type=int, default=8097, help='visdom port of the web display') self.parser.add_argument('--display_id', type=int, default=0, help='window id of the web display') # for generator self.parser.add_argument('--use_modulated_conv', type=bool, default = True, help='if specified, use modulated conv layers in the decoder like in StyleGAN2') self.parser.add_argument('--conv_weight_norm', type=bool, default = True, help='if specified, use weight normalization in conv and linear layers like in progrssive growing of GANs') self.parser.add_argument('--id_enc_norm', type=str, default='pixel', help='instance, pixel normalization') self.parser.add_argument('--decoder_norm',type=str, default='pixel', choices=['pixel','none'], help='type of upsampling layers normalization') self.parser.add_argument('--n_adaptive_blocks', type=int, default=4, help='# of adaptive normalization blocks') self.parser.add_argument('--activation',type=str, default='lrelu', choices=['relu','lrelu'], help='type of generator activation layer') self.parser.add_argument('--normalize_mlp', type=bool, default = True, help='if specified, normalize the generator MLP inputs and outputs') self.parser.add_argument('--no_moving_avg', action='store_true', help='if specified, do not use moving average network') self.parser.add_argument('--use_resblk_pixel_norm', action='store_true', help='if specified, apply pixel norm on the resnet block outputs') self.parser.add_argument('--ngf', type=int, default=64, help='# of gen filters in first conv layer') self.parser.add_argument('--no_cond_noise', action='store_true', help='remove gaussian noise from latent age code') self.parser.add_argument('--gen_dim_per_style', type=int, default=50, help='per class dimension of adain generator style latent code') self.parser.add_argument('--n_downsample', type=int, default=2, help='number of downsampling layers in generator') self.parser.add_argument('--verbose', action='store_true', default = True, help='toggles verbose') self.initialized = True def parse(self, save=True): if not self.initialized: self.initialize() try: self.opt = self.parser.parse_args() except: # solves argparse error in google colab self.opt = self.parser.parse_args(args=[]) self.opt.isTrain = self.isTrain # train or test str_ids = self.opt.gpu_ids.split(',') self.opt.gpu_ids = [] for str_id in str_ids: id = int(str_id) if id >= 0: self.opt.gpu_ids.append(id) # set gpu ids if len(self.opt.gpu_ids) > 0: torch.cuda.set_device(self.opt.gpu_ids[0]) # set class specific sort order if self.opt.sort_order is not None: order = self.opt.sort_order.split(',') self.opt.sort_order = [] for currName in order: self.opt.sort_order += [currName] # set decay schedule if self.isTrain and self.opt.decay_epochs is not None: decay_epochs = self.opt.decay_epochs.split(',') self.opt.decay_epochs = [] for curr_epoch in decay_epochs: self.opt.decay_epochs += [int(curr_epoch)] # create full image paths in traverse/deploy mode if (not self.isTrain) and (self.opt.traverse or self.opt.deploy): with open(self.opt.image_path_file,'r') as f: # temp_paths = f.read().splitlines() self.opt.image_path_list = f.read().splitlines() args = vars(self.opt) print('------------ Options -------------') for k, v in sorted(args.items()): print('%s: %s' % (str(k), str(v))) print('-------------- End ----------------') # save to the disk expr_dir = os.path.join(self.opt.checkpoints_dir, self.opt.name) util.mkdirs(expr_dir) if save:# and not self.opt.continue_train: file_name = os.path.join(expr_dir, 'opt.txt') with open(file_name, 'wt') as opt_file: opt_file.write('------------ Options -------------\n') for k, v in sorted(args.items()): opt_file.write('%s: %s\n' % (str(k), str(v))) opt_file.write('-------------- End ----------------\n') return self.opt