config.py

#######################
#author: Shiming Chen
#FREE
#######################
import argparse

parser = argparse.ArgumentParser()
parser.add_argument('--dataset', default='FLO', help='FLO')
parser.add_argument('--dataroot', default='data', help='path to dataset')
parser.add_argument('--image_embedding', default='res101')
parser.add_argument('--class_embedding', default='att')
parser.add_argument('--syn_num', type=int, default=100, help='number features to generate per class')
parser.add_argument('--gzsl', action='store_true', default=False, help='enable generalized zero-shot learning')
parser.add_argument('--preprocessing', action='store_true', default=False, help='enbale MinMaxScaler on visual features')
parser.add_argument('--standardization', action='store_true', default=False)
parser.add_argument('--workers', type=int, help='number of data loading workers', default=8)
parser.add_argument('--batch_size', type=int, default=64, help='input batch size')
parser.add_argument('--resSize', type=int, default=2048, help='size of visual features')
parser.add_argument('--attSize', type=int, default=1024, help='size of semantic features')
parser.add_argument('--nz', type=int, default=312, help='size of the latent z vector')
parser.add_argument('--ngh', type=int, default=4096, help='size of the hidden units in generator')
parser.add_argument('--ndh', type=int, default=1024, help='size of the hidden units in discriminator')
parser.add_argument('--nepoch', type=int, default=2000, help='number of epochs to train for')
parser.add_argument('--cls_nepoch', type=int, default=2000, help='number of epochs to train for classifier')
parser.add_argument('--critic_iter', type=int, default=5, help='critic iteration, following WGAN-GP')
parser.add_argument('--lambda1', type=float, default=10, help='gradient penalty regularizer, following WGAN-GP')
parser.add_argument('--lambda2', type=float, default=10, help='gradient penalty regularizer, following WGAN-GP')
parser.add_argument('--lr', type=float, default=0.001, help='learning rate to train GANs ')
parser.add_argument('--feed_lr', type=float, default=0.0001, help='learning rate to train GANs ')
parser.add_argument('--dec_lr', type=float, default=0.0001, help='learning rate to train GANs ')
parser.add_argument('--classifier_lr', type=float, default=0.001, help='learning rate to train softmax classifier')
parser.add_argument('--beta1', type=float, default=0.5, help='beta1 for adam. default=0.5')
parser.add_argument('--cuda', action='store_true', default=True, help='enables cuda')
parser.add_argument('--encoded_noise', action='store_true', default=False, help='enables validation mode')
parser.add_argument('--manualSeed', type=int, help='manual seed')
parser.add_argument('--nclass_all', type=int, default=200, help='number of all classes')
parser.add_argument('--validation', action='store_true', default=False, help='enables validation mode')
parser.add_argument("--encoder_layer_sizes", type=list, default=[8192, 4096])
parser.add_argument("--decoder_layer_sizes", type=list, default=[4096, 8192])
parser.add_argument('--gammaD', type=int, default=1000, help='weight on the W-GAN loss')
parser.add_argument('--gammaG', type=int, default=1000, help='weight on the W-GAN loss')
parser.add_argument('--gammaG_D2', type=int, default=1000, help='weight on the W-GAN loss')
parser.add_argument('--gammaD2', type=int, default=1000, help='weight on the W-GAN loss')
parser.add_argument("--latent_size", type=int, default=312)
parser.add_argument("--conditional", action='store_true',default=True)
###

parser.add_argument('--a1', type=float, default=1.0)
parser.add_argument('--a2', type=float, default=1.0)
parser.add_argument('--recons_weight', type=float, default=0.01, help='recons_weight for decoder')
parser.add_argument('--loop', type=int, default=2)
parser.add_argument('--freeze_dec', action='store_true', default=False, help='Freeze Decoder for fake samples')
#############################################################
parser.add_argument('--result_root', type=str, default='/home/shimingchen/ZSL/FREE/zero-shot-images/result', help='root path for saving checkpoint')
parser.add_argument('--center_margin', type=float, default=150, help='the margin in the center loss')
parser.add_argument('--center_weight', type=float, default=0.5, help='the weight for the center loss')
parser.add_argument('--incenter_weight', type=float, default=0.5, help='the weight for the center loss')
parser.add_argument('--cls_weight', type=float, default=0.2, help='weight of the classification loss')
parser.add_argument('--nclass_seen', type=int, default=150, help='number of seen classes')
parser.add_argument('--latensize', type=int, default=2048, help='size of semantic features')
parser.add_argument('--i_c', type=float, default=0.1, help='information constrain')
parser.add_argument('--lr_dec', action='store_true', default=False, help='enable lr decay or not')
parser.add_argument('--lr_dec_ep', type=int, default=1, help='lr decay for every 100 epoch')

#################################################################
opt = parser.parse_args()
opt.lambda2 = opt.lambda1
opt.encoder_layer_sizes[0] = opt.resSize
opt.decoder_layer_sizes[-1] = opt.resSize
opt.latent_size = opt.attSize