Exemplo n.º 1
0
def build_models():
    # build model ############################################################
    text_encoder = RNN_ENCODER(dataset.n_words, nhidden=cfg.TEXT.EMBEDDING_DIM)
    image_encoder = CNN_ENCODER(cfg.TEXT.EMBEDDING_DIM)
    labels = Variable(torch.LongTensor(range(batch_size)))
    start_epoch = 0
    if cfg.TRAIN.NET_E != '':
        state_dict = torch.load(cfg.TRAIN.NET_E)
        text_encoder.load_state_dict(state_dict)
        print('Load ', cfg.TRAIN.NET_E)
        #
        name = cfg.TRAIN.NET_E.replace('text_encoder', 'image_encoder')
        state_dict = torch.load(name)
        image_encoder.load_state_dict(state_dict)
        print('Load ', name)

        istart = cfg.TRAIN.NET_E.rfind('_') + 8
        iend = cfg.TRAIN.NET_E.rfind('.')
        start_epoch = cfg.TRAIN.NET_E[istart:iend]
        start_epoch = int(start_epoch) + 1
        print('start_epoch', start_epoch)
    if cfg.CUDA:
        text_encoder = text_encoder.cuda()
        image_encoder = image_encoder.cuda()
        labels = labels.cuda()

    return text_encoder, image_encoder, labels, start_epoch
Exemplo n.º 2
0
    def __init__(self, opt):
        """Initialize the pix2pix class.

        Parameters:
            opt (Option class)-- stores all the experiment flags; needs to be a subclass of BaseOptions
        """
        BaseModel.__init__(self, opt)
        self.batch_size = opt.batch_size
        self.img_size = opt.crop_size
        # specify the training losses you want to print out. The training/test scripts will call <BaseModel.get_current_losses>
        self.loss_names = ['G_GAN', 'G_L1', 'D_real', 'D_fake']
        # specify the images you want to save/display. The training/test scripts will call <BaseModel.get_current_visuals>
        self.visual_names = ['real_A', 'fake_B', 'real_B']
        # specify the models you want to save to the disk. The training/test scripts will call <BaseModel.save_networks> and <BaseModel.load_networks>
        if self.isTrain:
            self.model_names = ['G', 'D']
        else:  # during test time, only load G
            self.model_names = ['G']
        # define networks (both generator and discriminator)
        self.text_encoder = RNN_ENCODER(opt.text_words_num,
                                        nhidden=opt.text_embedding_dim).to(
                                            self.device)
        state_dict = torch.load(opt.text_encoder,
                                map_location=lambda storage, loc: storage)
        self.text_encoder.load_state_dict(state_dict)
        for p in self.text_encoder.parameters():
            p.requires_grad = False
        print('Load text encoder from:', opt.text_encoder)
        self.text_encoder.eval()

        self.netG = networks.define_G(opt.input_nc, opt.output_nc, opt.ngf,
                                      opt.netG, opt.norm, not opt.no_dropout,
                                      opt.init_type, opt.init_gain,
                                      self.gpu_ids)

        if self.isTrain:  # define a discriminator; conditional GANs need to take both input and output images; Therefore, #channels for D is input_nc + output_nc
            self.netD = networks.define_D(opt.input_nc + opt.output_nc,
                                          opt.ndf, opt.netD, opt.n_layers_D,
                                          opt.norm, opt.init_type,
                                          opt.init_gain, self.gpu_ids)

        if self.isTrain:
            # define loss functions
            self.criterionGAN = networks.GANLoss(opt.gan_mode).to(self.device)
            self.criterionL1 = torch.nn.L1Loss()
            # initialize optimizers; schedulers will be automatically created by function <BaseModel.setup>.
            self.optimizer_G = torch.optim.Adam(self.netG.parameters(),
                                                lr=opt.lr,
                                                betas=(opt.beta1, 0.999))
            self.optimizer_D = torch.optim.Adam(self.netD.parameters(),
                                                lr=opt.lr,
                                                betas=(opt.beta1, 0.999))
            self.optimizers.append(self.optimizer_G)
            self.optimizers.append(self.optimizer_D)
Exemplo n.º 3
0
def models(word_len):
    cfg_from_file('../flask-server/AttnGAN/code/cfg/eval_plans2.yaml')
    text_encoder = cache.get('text_encoder')
    if text_encoder is None:
        #print("text_encoder not cached")
        text_encoder = RNN_ENCODER(word_len, nhidden=cfg.TEXT.EMBEDDING_DIM)
        state_dict = torch.load(cfg.TRAIN.NET_E,
                                map_location=lambda storage, loc: storage)
        text_encoder.load_state_dict(state_dict)
        if cfg.CUDA:
            text_encoder.cuda()
        text_encoder.eval()
        cache.set('text_encoder', text_encoder, timeout=60 * 60 * 24)

    netG = cache.get('netG')
    if netG is None:
        #print("netG not cached")
        netG = G_NET()
        state_dict = torch.load(cfg.TRAIN.NET_G,
                                map_location=lambda storage, loc: storage)
        netG.load_state_dict(state_dict)
        if cfg.CUDA:
            netG.cuda()
        netG.eval()
        cache.set('netG', netG, timeout=60 * 60 * 24)

    return text_encoder, netG
Exemplo n.º 4
0
    def __init__(self, output_dir, data_loader, n_words, ixtoword):
        if cfg.TRAIN.FLAG:
            self.model_dir = os.path.join(output_dir, 'Model')
            self.image_dir = os.path.join(output_dir, 'Image')
            mkdir_p(self.model_dir)
            mkdir_p(self.image_dir)

        torch.cuda.set_device(cfg.GPU_ID)
        cudnn.benchmark = True

        self.batch_size = cfg.TRAIN.BATCH_SIZE
        self.max_epoch = cfg.TRAIN.MAX_EPOCH
        self.snapshot_interval = cfg.TRAIN.SNAPSHOT_INTERVAL

        self.n_words = n_words
        self.ixtoword = ixtoword
        self.data_loader = data_loader
        self.num_batches = len(self.data_loader)
        
        
        # Build and load the generator
        self.text_encoder = \
            RNN_ENCODER(self.n_words, nhidden=cfg.TEXT.EMBEDDING_DIM)
        state_dict = \
            torch.load(cfg.TRAIN.NET_E, map_location=lambda storage, loc: storage)
        self.text_encoder.load_state_dict(state_dict)
        print('Load text encoder from:', cfg.TRAIN.NET_E)
        self.text_encoder = self.text_encoder.cuda()
        self.text_encoder.eval()

        # the path to save generated images
        if cfg.GAN.B_DCGAN:
            self.netG = G_DCGAN()
        else:
            self.netG = G_NET()
        s_tmp = cfg.TRAIN.NET_G[:cfg.TRAIN.NET_G.rfind('.pth')]
        model_dir = cfg.TRAIN.NET_G
        state_dict = \
            torch.load(model_dir, map_location=lambda storage, loc: storage)
        self.netG.load_state_dict(state_dict)
        print('Load G from: ', model_dir)
        self.netG.cuda()
        self.netG.eval()                        
Exemplo n.º 5
0
class Pix2PixModel(BaseModel):
    """ This class implements the pix2pix model, for learning a mapping from input images to output images given paired data.

    The model training requires '--dataset_mode aligned' dataset.
    By default, it uses a '--netG unet256' U-Net generator,
    a '--netD basic' discriminator (PatchGAN),
    and a '--gan_mode' vanilla GAN loss (the cross-entropy objective used in the orignal GAN paper).

    pix2pix paper: https://arxiv.org/pdf/1611.07004.pdf
    """
    @staticmethod
    def modify_commandline_options(parser, is_train=True):
        """Add new dataset-specific options, and rewrite default values for existing options.

        Parameters:
            parser          -- original option parser
            is_train (bool) -- whether training phase or test phase. You can use this flag to add training-specific or test-specific options.

        Returns:
            the modified parser.

        For pix2pix, we do not use image buffer
        The training objective is: GAN Loss + lambda_L1 * ||G(A)-B||_1
        By default, we use vanilla GAN loss, UNet with batchnorm, and aligned datasets.
        """
        # changing the default values to match the pix2pix paper (https://phillipi.github.io/pix2pix/)
        parser.set_defaults(norm='batch',
                            netG='unet_256',
                            dataset_mode='aligned')
        if is_train:
            parser.set_defaults(pool_size=0, gan_mode='vanilla')
            parser.add_argument('--lambda_L1',
                                type=float,
                                default=100.0,
                                help='weight for L1 loss')

        return parser

    def __init__(self, opt):
        """Initialize the pix2pix class.

        Parameters:
            opt (Option class)-- stores all the experiment flags; needs to be a subclass of BaseOptions
        """
        BaseModel.__init__(self, opt)
        self.batch_size = opt.batch_size
        self.img_size = opt.crop_size
        # specify the training losses you want to print out. The training/test scripts will call <BaseModel.get_current_losses>
        self.loss_names = ['G_GAN', 'G_L1', 'D_real', 'D_fake']
        # specify the images you want to save/display. The training/test scripts will call <BaseModel.get_current_visuals>
        self.visual_names = ['real_A', 'fake_B', 'real_B']
        # specify the models you want to save to the disk. The training/test scripts will call <BaseModel.save_networks> and <BaseModel.load_networks>
        if self.isTrain:
            self.model_names = ['G', 'D']
        else:  # during test time, only load G
            self.model_names = ['G']
        # define networks (both generator and discriminator)
        self.text_encoder = RNN_ENCODER(opt.text_words_num,
                                        nhidden=opt.text_embedding_dim).to(
                                            self.device)
        state_dict = torch.load(opt.text_encoder,
                                map_location=lambda storage, loc: storage)
        self.text_encoder.load_state_dict(state_dict)
        for p in self.text_encoder.parameters():
            p.requires_grad = False
        print('Load text encoder from:', opt.text_encoder)
        self.text_encoder.eval()

        self.netG = networks.define_G(opt.input_nc, opt.output_nc, opt.ngf,
                                      opt.netG, opt.norm, not opt.no_dropout,
                                      opt.init_type, opt.init_gain,
                                      self.gpu_ids)

        if self.isTrain:  # define a discriminator; conditional GANs need to take both input and output images; Therefore, #channels for D is input_nc + output_nc
            self.netD = networks.define_D(opt.input_nc + opt.output_nc,
                                          opt.ndf, opt.netD, opt.n_layers_D,
                                          opt.norm, opt.init_type,
                                          opt.init_gain, self.gpu_ids)

        if self.isTrain:
            # define loss functions
            self.criterionGAN = networks.GANLoss(opt.gan_mode).to(self.device)
            self.criterionL1 = torch.nn.L1Loss()
            # initialize optimizers; schedulers will be automatically created by function <BaseModel.setup>.
            self.optimizer_G = torch.optim.Adam(self.netG.parameters(),
                                                lr=opt.lr,
                                                betas=(opt.beta1, 0.999))
            self.optimizer_D = torch.optim.Adam(self.netD.parameters(),
                                                lr=opt.lr,
                                                betas=(opt.beta1, 0.999))
            self.optimizers.append(self.optimizer_G)
            self.optimizers.append(self.optimizer_D)

    def set_input(self, input):
        """Unpack input data from the dataloader and perform necessary pre-processing steps.

        Parameters:
            input (dict): include the data itself and its metadata information.

        The option 'direction' can be used to swap images in domain A and domain B.
        """
        AtoB = self.opt.direction == 'AtoB'
        self.caption_len, sort_idx = input["caption_len"].sort(descending=True)
        self.real_A = input['A' if AtoB else 'B'][[sort_idx]].to(self.device)
        self.real_B = input['B' if AtoB else 'A'][[sort_idx]].to(self.device)
        self.caption = input["caption"][[sort_idx]].to(self.device)
        self.image_paths = list(
            np.array(
                input['A_paths' if AtoB else 'B_paths'])[[sort_idx.tolist()]])

        # Encode text
        hidden = self.text_encoder.init_hidden(self.batch_size)
        _, self.sent_emb = self.text_encoder(
            self.caption, self.caption_len,
            hidden)  # sent_emb: [batch_size(1), sent_dim(128)]
        self.tiled_sentence = self.sent_emb.unsqueeze(2).unsqueeze(3).repeat(
            1, 1, self.img_size, self.img_size)

    def forward(self):
        """Run forward pass; called by both functions <optimize_parameters> and <test>."""
        real_A = torch.cat(
            (self.real_A, self.tiled_sentence), 1
        )  # real_A: [batch_size(1), 3+sent_dim(128), crop_size(256), crop_size(256)]
        self.fake_B = self.netG(real_A)  # G(A)

    def backward_D(self):
        """Calculate GAN loss for the discriminator"""
        # Fake; stop backprop to the generator by detaching fake_B
        fake_AB = torch.cat(
            (self.real_A, self.fake_B, self.tiled_sentence), 1
        )  # we use conditional GANs; we need to feed both input and output to the discriminator
        pred_fake = self.netD(fake_AB.detach())
        self.loss_D_fake = self.criterionGAN(pred_fake, False)
        # Real
        real_AB = torch.cat((self.real_A, self.real_B, self.tiled_sentence), 1)
        pred_real = self.netD(real_AB)
        self.loss_D_real = self.criterionGAN(pred_real, True)
        # combine loss and calculate gradients
        self.loss_D = (self.loss_D_fake + self.loss_D_real) * 0.5
        self.loss_D.backward()

    def backward_G(self):
        """Calculate GAN and L1 loss for the generator"""
        # First, G(A) should fake the discriminator
        fake_AB = torch.cat((self.real_A, self.fake_B, self.tiled_sentence), 1)
        pred_fake = self.netD(fake_AB)
        self.loss_G_GAN = self.criterionGAN(pred_fake, True)
        # Second, G(A) = B
        self.loss_G_L1 = self.criterionL1(self.fake_B,
                                          self.real_B) * self.opt.lambda_L1
        # combine loss and calculate gradients
        self.loss_G = self.loss_G_GAN + self.loss_G_L1
        self.loss_G.backward()

    def optimize_parameters(self):
        self.forward()  # compute fake images: G(A)
        # update D
        self.set_requires_grad(self.netD, True)  # enable backprop for D
        self.optimizer_D.zero_grad()  # set D's gradients to zero
        self.backward_D()  # calculate gradients for D
        self.optimizer_D.step()  # update D's weights
        # update G
        self.set_requires_grad(
            self.netD, False)  # D requires no gradients when optimizing G
        self.optimizer_G.zero_grad()  # set G's gradients to zero
        self.backward_G()  # calculate graidents for G
        self.optimizer_G.step()  # udpate G's weights
Exemplo n.º 6
0
    def build_models(self):
        # ###################encoders######################################## #
        if cfg.TRAIN.NET_E == '':
            print('Error: no pretrained text-image encoders')
            return

        image_encoder = CNN_ENCODER(cfg.TEXT.EMBEDDING_DIM)
        img_encoder_path = cfg.TRAIN.NET_E.replace('text_encoder', 'image_encoder')
        state_dict = \
            torch.load(img_encoder_path, map_location=lambda storage, loc: storage)
        image_encoder.load_state_dict(state_dict)
        for p in image_encoder.parameters():
            p.requires_grad = False
        print('Load image encoder from:', img_encoder_path)
        image_encoder.eval()

        text_encoder = \
            RNN_ENCODER(self.n_words, nhidden=cfg.TEXT.EMBEDDING_DIM)
        state_dict = \
            torch.load(cfg.TRAIN.NET_E,
                       map_location=lambda storage, loc: storage)
        text_encoder.load_state_dict(state_dict)
        for p in text_encoder.parameters():
            p.requires_grad = False
        print('Load text encoder from:', cfg.TRAIN.NET_E)
        text_encoder.eval()

        # #######################generator and discriminators############## #
        netsD = []
        if cfg.GAN.B_DCGAN:
            if cfg.TREE.BRANCH_NUM ==1:
                from model import D_NET64 as D_NET
            elif cfg.TREE.BRANCH_NUM == 2:
                from model import D_NET128 as D_NET
            else:  # cfg.TREE.BRANCH_NUM == 3:
                from model import D_NET256 as D_NET
            # TODO: elif cfg.TREE.BRANCH_NUM > 3:
            netG = G_DCGAN()
            netsD = [D_NET(b_jcu=False)]
        else:
            from model import D_NET64, D_NET128, D_NET256
            netG = G_NET()
            if cfg.TREE.BRANCH_NUM > 0:
                netsD.append(D_NET64())
            if cfg.TREE.BRANCH_NUM > 1:
                netsD.append(D_NET128())
            if cfg.TREE.BRANCH_NUM > 2:
                netsD.append(D_NET256())
            # TODO: if cfg.TREE.BRANCH_NUM > 3:
        netG.apply(weights_init)
        # print(netG)
        for i in range(len(netsD)):
            netsD[i].apply(weights_init)
            # print(netsD[i])
        print('# of netsD', len(netsD))
        #
        epoch = 0
        if cfg.TRAIN.NET_G != '':
            state_dict = \
                torch.load(cfg.TRAIN.NET_G, map_location=lambda storage, loc: storage)
            netG.load_state_dict(state_dict)
            print('Load G from: ', cfg.TRAIN.NET_G)
            istart = cfg.TRAIN.NET_G.rfind('_') + 1
            iend = cfg.TRAIN.NET_G.rfind('.')
            epoch = cfg.TRAIN.NET_G[istart:iend]
            epoch = int(epoch) + 1
            if cfg.TRAIN.B_NET_D:
                Gname = cfg.TRAIN.NET_G
                for i in range(len(netsD)):
                    s_tmp = Gname[:Gname.rfind('/')]
                    Dname = '%s/netD%d.pth' % (s_tmp, i)
                    print('Load D from: ', Dname)
                    state_dict = \
                        torch.load(Dname, map_location=lambda storage, loc: storage)
                    netsD[i].load_state_dict(state_dict)
        # ########################################################### #
        if cfg.CUDA:
            text_encoder = text_encoder.cuda()
            image_encoder = image_encoder.cuda()
            netG.cuda()
            for i in range(len(netsD)):
                netsD[i].cuda()
        return [text_encoder, image_encoder, netG, netsD, epoch]
Exemplo n.º 7
0
    def sampling(self, split_dir):
        if cfg.TRAIN.NET_G == '':
            print('Error: the path for morels is not found!')
        else:
            if split_dir == 'test':
                split_dir = 'valid'
            # Build and load the generator
            if cfg.GAN.B_DCGAN:
                netG = G_DCGAN()
            else:
                netG = G_NET()
            netG.apply(weights_init)
            netG.cuda()
            netG.eval()
            #
            text_encoder = RNN_ENCODER(self.n_words, nhidden=cfg.TEXT.EMBEDDING_DIM)
            state_dict = \
                torch.load(cfg.TRAIN.NET_E, map_location=lambda storage, loc: storage)
            text_encoder.load_state_dict(state_dict)
            print('Load text encoder from:', cfg.TRAIN.NET_E)
            text_encoder = text_encoder.cuda()
            text_encoder.eval()

            batch_size = self.batch_size
            nz = cfg.GAN.Z_DIM
            noise = Variable(torch.FloatTensor(batch_size, nz), volatile=True)
            noise = noise.cuda()

            model_dir = cfg.TRAIN.NET_G
            state_dict = \
                torch.load(model_dir, map_location=lambda storage, loc: storage)
            # state_dict = torch.load(cfg.TRAIN.NET_G)
            netG.load_state_dict(state_dict)
            print('Load G from: ', model_dir)

            # the path to save generated images
            s_tmp = model_dir[:model_dir.rfind('.pth')]
            save_dir = '%s/%s' % (s_tmp, split_dir)
            mkdir_p(save_dir)

            cnt = 0

            for _ in range(1):  # (cfg.TEXT.CAPTIONS_PER_IMAGE):
                for step, data in enumerate(self.data_loader, 0):
                    cnt += batch_size
                    if step % 100 == 0:
                        print('step: ', step)
                    # if step > 50:
                    #     break

                    imgs, captions, cap_lens, class_ids, keys = prepare_data(data)

                    hidden = text_encoder.init_hidden(batch_size)
                    # words_embs: batch_size x nef x seq_len
                    # sent_emb: batch_size x nef
                    words_embs, sent_emb = text_encoder(captions, cap_lens, hidden)
                    words_embs, sent_emb = words_embs.detach(), sent_emb.detach()
                    mask = (captions == 0)
                    num_words = words_embs.size(2)
                    if mask.size(1) > num_words:
                        mask = mask[:, :num_words]

                    #######################################################
                    # (2) Generate fake images
                    ######################################################
                    noise.data.normal_(0, 1)
                    fake_imgs, _, _, _ = netG(noise, sent_emb, words_embs, mask)
                    for j in range(batch_size):
                        s_tmp = '%s/single/%s' % (save_dir, keys[j])
                        folder = s_tmp[:s_tmp.rfind('/')]
                        if not os.path.isdir(folder):
                            print('Make a new folder: ', folder)
                            mkdir_p(folder)
                        k = -1
                        # for k in range(len(fake_imgs)):
                        im = fake_imgs[k][j].data.cpu().numpy()
                        # [-1, 1] --> [0, 255]
                        im = (im + 1.0) * 127.5
                        im = im.astype(np.uint8)
                        im = np.transpose(im, (1, 2, 0))
                        im = Image.fromarray(im)
                        fullpath = '%s_s%d.png' % (s_tmp, k)
                        im.save(fullpath)
Exemplo n.º 8
0
class condGANTrainer(object):
    def __init__(self, output_dir, data_loader, n_words, ixtoword):
        if cfg.TRAIN.FLAG:
            self.model_dir = os.path.join(output_dir, 'Model')
            self.image_dir = os.path.join(output_dir, 'Image')
            mkdir_p(self.model_dir)
            mkdir_p(self.image_dir)

        torch.cuda.set_device(cfg.GPU_ID)
        cudnn.benchmark = True

        self.batch_size = cfg.TRAIN.BATCH_SIZE
        self.max_epoch = cfg.TRAIN.MAX_EPOCH
        self.snapshot_interval = cfg.TRAIN.SNAPSHOT_INTERVAL

        self.n_words = n_words
        self.ixtoword = ixtoword
        self.data_loader = data_loader
        self.num_batches = len(self.data_loader)
        
        
        # Build and load the generator
        self.text_encoder = \
            RNN_ENCODER(self.n_words, nhidden=cfg.TEXT.EMBEDDING_DIM)
        state_dict = \
            torch.load(cfg.TRAIN.NET_E, map_location=lambda storage, loc: storage)
        self.text_encoder.load_state_dict(state_dict)
        print('Load text encoder from:', cfg.TRAIN.NET_E)
        self.text_encoder = self.text_encoder.cuda()
        self.text_encoder.eval()

        # the path to save generated images
        if cfg.GAN.B_DCGAN:
            self.netG = G_DCGAN()
        else:
            self.netG = G_NET()
        s_tmp = cfg.TRAIN.NET_G[:cfg.TRAIN.NET_G.rfind('.pth')]
        model_dir = cfg.TRAIN.NET_G
        state_dict = \
            torch.load(model_dir, map_location=lambda storage, loc: storage)
        self.netG.load_state_dict(state_dict)
        print('Load G from: ', model_dir)
        self.netG.cuda()
        self.netG.eval()                        
               

    def build_models(self):
        # ###################encoders######################################## #
        if cfg.TRAIN.NET_E == '':
            print('Error: no pretrained text-image encoders')
            return

        image_encoder = CNN_ENCODER(cfg.TEXT.EMBEDDING_DIM)
        img_encoder_path = cfg.TRAIN.NET_E.replace('text_encoder', 'image_encoder')
        state_dict = \
            torch.load(img_encoder_path, map_location=lambda storage, loc: storage)
        image_encoder.load_state_dict(state_dict)
        for p in image_encoder.parameters():
            p.requires_grad = False
        print('Load image encoder from:', img_encoder_path)
        image_encoder.eval()

        text_encoder = \
            RNN_ENCODER(self.n_words, nhidden=cfg.TEXT.EMBEDDING_DIM)
        state_dict = \
            torch.load(cfg.TRAIN.NET_E,
                       map_location=lambda storage, loc: storage)
        text_encoder.load_state_dict(state_dict)
        for p in text_encoder.parameters():
            p.requires_grad = False
        print('Load text encoder from:', cfg.TRAIN.NET_E)
        text_encoder.eval()

        # #######################generator and discriminators############## #
        netsD = []
        if cfg.GAN.B_DCGAN:
            if cfg.TREE.BRANCH_NUM ==1:
                from model import D_NET64 as D_NET
            elif cfg.TREE.BRANCH_NUM == 2:
                from model import D_NET128 as D_NET
            else:  # cfg.TREE.BRANCH_NUM == 3:
                from model import D_NET256 as D_NET
            # TODO: elif cfg.TREE.BRANCH_NUM > 3:
            netG = G_DCGAN()
            netsD = [D_NET(b_jcu=False)]
        else:
            from model import D_NET64, D_NET128, D_NET256
            netG = G_NET()
            if cfg.TREE.BRANCH_NUM > 0:
                netsD.append(D_NET64())
            if cfg.TREE.BRANCH_NUM > 1:
                netsD.append(D_NET128())
            if cfg.TREE.BRANCH_NUM > 2:
                netsD.append(D_NET256())
            # TODO: if cfg.TREE.BRANCH_NUM > 3:
        netG.apply(weights_init)
        # print(netG)
        for i in range(len(netsD)):
            netsD[i].apply(weights_init)
            # print(netsD[i])
        print('# of netsD', len(netsD))
        #
        epoch = 0
        if cfg.TRAIN.NET_G != '':
            state_dict = \
                torch.load(cfg.TRAIN.NET_G, map_location=lambda storage, loc: storage)
            netG.load_state_dict(state_dict)
            print('Load G from: ', cfg.TRAIN.NET_G)
            istart = cfg.TRAIN.NET_G.rfind('_') + 1
            iend = cfg.TRAIN.NET_G.rfind('.')
            epoch = cfg.TRAIN.NET_G[istart:iend]
            epoch = int(epoch) + 1
            if cfg.TRAIN.B_NET_D:
                Gname = cfg.TRAIN.NET_G
                for i in range(len(netsD)):
                    s_tmp = Gname[:Gname.rfind('/')]
                    Dname = '%s/netD%d.pth' % (s_tmp, i)
                    print('Load D from: ', Dname)
                    state_dict = \
                        torch.load(Dname, map_location=lambda storage, loc: storage)
                    netsD[i].load_state_dict(state_dict)
        # ########################################################### #
        if cfg.CUDA:
            text_encoder = text_encoder.cuda()
            image_encoder = image_encoder.cuda()
            netG.cuda()
            for i in range(len(netsD)):
                netsD[i].cuda()
        return [text_encoder, image_encoder, netG, netsD, epoch]

    def define_optimizers(self, netG, netsD):
        optimizersD = []
        num_Ds = len(netsD)
        for i in range(num_Ds):
            opt = optim.Adam(netsD[i].parameters(),
                             lr=cfg.TRAIN.DISCRIMINATOR_LR,
                             betas=(0.5, 0.999))
            optimizersD.append(opt)

        optimizerG = optim.Adam(netG.parameters(),
                                lr=cfg.TRAIN.GENERATOR_LR,
                                betas=(0.5, 0.999))

        return optimizerG, optimizersD

    def prepare_labels(self):
        batch_size = self.batch_size
        real_labels = Variable(torch.FloatTensor(batch_size).fill_(1))
        fake_labels = Variable(torch.FloatTensor(batch_size).fill_(0))
        match_labels = Variable(torch.LongTensor(range(batch_size)))
        if cfg.CUDA:
            real_labels = real_labels.cuda()
            fake_labels = fake_labels.cuda()
            match_labels = match_labels.cuda()

        return real_labels, fake_labels, match_labels

    def save_model(self, netG, avg_param_G, netsD, epoch):
        backup_para = copy_G_params(netG)
        load_params(netG, avg_param_G)
        torch.save(netG.state_dict(),
            '%s/netG_epoch_%d.pth' % (self.model_dir, epoch))
        load_params(netG, backup_para)
        #
        for i in range(len(netsD)):
            netD = netsD[i]
            torch.save(netD.state_dict(),
                '%s/netD%d.pth' % (self.model_dir, i))
        print('Save G/Ds models.')

    def set_requires_grad_value(self, models_list, brequires):
        for i in range(len(models_list)):
            for p in models_list[i].parameters():
                p.requires_grad = brequires

    def save_img_results(self, netG, noise, sent_emb, words_embs, mask,
                         image_encoder, captions, cap_lens,
                         gen_iterations, name='current'):
        # Save images
        fake_imgs, attention_maps, _, _ = netG(noise, sent_emb, words_embs, mask)
        for i in range(len(attention_maps)):
            if len(fake_imgs) > 1:
                img = fake_imgs[i + 1].detach().cpu()
                lr_img = fake_imgs[i].detach().cpu()
            else:
                img = fake_imgs[0].detach().cpu()
                lr_img = None
            attn_maps = attention_maps[i]
            att_sze = attn_maps.size(2)
            img_set, _ = \
                build_super_images(img, captions, self.ixtoword,
                                   attn_maps, att_sze, lr_imgs=lr_img)
            if img_set is not None:
                im = Image.fromarray(img_set)
                fullpath = '%s/G_%s_%d_%d.png'\
                    % (self.image_dir, name, gen_iterations, i)
                im.save(fullpath)

        # for i in range(len(netsD)):
        i = -1
        img = fake_imgs[i].detach()
        region_features, _ = image_encoder(img)
        att_sze = region_features.size(2)
        _, _, att_maps = words_loss(region_features.detach(),
                                    words_embs.detach(),
                                    None, cap_lens,
                                    None, self.batch_size)
        img_set, _ = \
            build_super_images(fake_imgs[i].detach().cpu(),
                               captions, self.ixtoword, att_maps, att_sze)
        if img_set is not None:
            im = Image.fromarray(img_set)
            fullpath = '%s/D_%s_%d.png'\
                % (self.image_dir, name, gen_iterations)
            im.save(fullpath)

    def train(self):
        text_encoder, image_encoder, netG, netsD, start_epoch = self.build_models()
        avg_param_G = copy_G_params(netG)
        optimizerG, optimizersD = self.define_optimizers(netG, netsD)
        real_labels, fake_labels, match_labels = self.prepare_labels()

        batch_size = self.batch_size
        nz = cfg.GAN.Z_DIM
        noise = Variable(torch.FloatTensor(batch_size, nz))
        fixed_noise = Variable(torch.FloatTensor(batch_size, nz).normal_(0, 1))
        if cfg.CUDA:
            noise, fixed_noise = noise.cuda(), fixed_noise.cuda()

        gen_iterations = 0
        # gen_iterations = start_epoch * self.num_batches
        for epoch in range(start_epoch, self.max_epoch):
            start_t = time.time()

            data_iter = iter(self.data_loader)
            step = 0
            while step < self.num_batches:
                # reset requires_grad to be trainable for all Ds
                # self.set_requires_grad_value(netsD, True)

                ######################################################
                # (1) Prepare training data and Compute text embeddings
                ######################################################
                data = data_iter.next()
                imgs, captions, cap_lens, class_ids, keys = prepare_data(data)

                hidden = text_encoder.init_hidden(batch_size)
                # words_embs: batch_size x nef x seq_len
                # sent_emb: batch_size x nef
                words_embs, sent_emb = text_encoder(captions, cap_lens, hidden)
                words_embs, sent_emb = words_embs.detach(), sent_emb.detach()
                mask = (captions == 0)
                num_words = words_embs.size(2)
                if mask.size(1) > num_words:
                    mask = mask[:, :num_words]

                #######################################################
                # (2) Generate fake images
                ######################################################
                noise.data.normal_(0, 1)
                fake_imgs, _, mu, logvar = netG(noise, sent_emb, words_embs, mask)

                #######################################################
                # (3) Update D network
                ######################################################
                errD_total = 0
                D_logs = ''
                for i in range(len(netsD)):
                    netsD[i].zero_grad()
                    errD = discriminator_loss(netsD[i], imgs[i], fake_imgs[i],
                                              sent_emb, real_labels, fake_labels)
                    # backward and update parameters
                    errD.backward()
                    optimizersD[i].step()
                    errD_total += errD
                    D_logs += 'errD%d: %.2f ' % (i, errD.data[0])

                #######################################################
                # (4) Update G network: maximize log(D(G(z)))
                ######################################################
                # compute total loss for training G
                step += 1
                gen_iterations += 1

                # do not need to compute gradient for Ds
                # self.set_requires_grad_value(netsD, False)
                netG.zero_grad()
                errG_total, G_logs = \
                    generator_loss(netsD, image_encoder, fake_imgs, real_labels,
                                   words_embs, sent_emb, match_labels, cap_lens, class_ids)
                kl_loss = KL_loss(mu, logvar)
                errG_total += kl_loss
                G_logs += 'kl_loss: %.2f ' % kl_loss.data[0]
                # backward and update parameters
                errG_total.backward()
                optimizerG.step()
                for p, avg_p in zip(netG.parameters(), avg_param_G):
                    avg_p.mul_(0.999).add_(0.001, p.data)

                if gen_iterations % 100 == 0:
                    print(D_logs + '\n' + G_logs)
                # save images
                if gen_iterations % 1000 == 0:
                    backup_para = copy_G_params(netG)
                    load_params(netG, avg_param_G)
                    self.save_img_results(netG, fixed_noise, sent_emb,
                                          words_embs, mask, image_encoder,
                                          captions, cap_lens, epoch, name='average')
                    load_params(netG, backup_para)
                    #
                    # self.save_img_results(netG, fixed_noise, sent_emb,
                    #                       words_embs, mask, image_encoder,
                    #                       captions, cap_lens,
                    #                       epoch, name='current')
            end_t = time.time()

            print('''[%d/%d][%d]
                  Loss_D: %.2f Loss_G: %.2f Time: %.2fs'''
                  % (epoch, self.max_epoch, self.num_batches,
                     errD_total.data[0], errG_total.data[0],
                     end_t - start_t))

            if epoch % cfg.TRAIN.SNAPSHOT_INTERVAL == 0:  # and epoch != 0:
                self.save_model(netG, avg_param_G, netsD, epoch)

        self.save_model(netG, avg_param_G, netsD, self.max_epoch)

    def save_singleimages(self, images, filenames, save_dir,
                          split_dir, sentenceID=0):
        for i in range(images.size(0)):
            s_tmp = '%s/single_samples/%s/%s' %\
                (save_dir, split_dir, filenames[i])
            folder = s_tmp[:s_tmp.rfind('/')]
            if not os.path.isdir(folder):
                print('Make a new folder: ', folder)
                mkdir_p(folder)

            fullpath = '%s_%d.jpg' % (s_tmp, sentenceID)
            # range from [-1, 1] to [0, 1]
            # img = (images[i] + 1.0) / 2
            img = images[i].add(1).div(2).mul(255).clamp(0, 255).byte()
            # range from [0, 1] to [0, 255]
            ndarr = img.permute(1, 2, 0).data.cpu().numpy()
            im = Image.fromarray(ndarr)
            im.save(fullpath)

    def sampling(self, split_dir):
        if cfg.TRAIN.NET_G == '':
            print('Error: the path for morels is not found!')
        else:
            if split_dir == 'test':
                split_dir = 'valid'
            # Build and load the generator
            if cfg.GAN.B_DCGAN:
                netG = G_DCGAN()
            else:
                netG = G_NET()
            netG.apply(weights_init)
            netG.cuda()
            netG.eval()
            #
            text_encoder = RNN_ENCODER(self.n_words, nhidden=cfg.TEXT.EMBEDDING_DIM)
            state_dict = \
                torch.load(cfg.TRAIN.NET_E, map_location=lambda storage, loc: storage)
            text_encoder.load_state_dict(state_dict)
            print('Load text encoder from:', cfg.TRAIN.NET_E)
            text_encoder = text_encoder.cuda()
            text_encoder.eval()

            batch_size = self.batch_size
            nz = cfg.GAN.Z_DIM
            noise = Variable(torch.FloatTensor(batch_size, nz), volatile=True)
            noise = noise.cuda()

            model_dir = cfg.TRAIN.NET_G
            state_dict = \
                torch.load(model_dir, map_location=lambda storage, loc: storage)
            # state_dict = torch.load(cfg.TRAIN.NET_G)
            netG.load_state_dict(state_dict)
            print('Load G from: ', model_dir)

            # the path to save generated images
            s_tmp = model_dir[:model_dir.rfind('.pth')]
            save_dir = '%s/%s' % (s_tmp, split_dir)
            mkdir_p(save_dir)

            cnt = 0

            for _ in range(1):  # (cfg.TEXT.CAPTIONS_PER_IMAGE):
                for step, data in enumerate(self.data_loader, 0):
                    cnt += batch_size
                    if step % 100 == 0:
                        print('step: ', step)
                    # if step > 50:
                    #     break

                    imgs, captions, cap_lens, class_ids, keys = prepare_data(data)

                    hidden = text_encoder.init_hidden(batch_size)
                    # words_embs: batch_size x nef x seq_len
                    # sent_emb: batch_size x nef
                    words_embs, sent_emb = text_encoder(captions, cap_lens, hidden)
                    words_embs, sent_emb = words_embs.detach(), sent_emb.detach()
                    mask = (captions == 0)
                    num_words = words_embs.size(2)
                    if mask.size(1) > num_words:
                        mask = mask[:, :num_words]

                    #######################################################
                    # (2) Generate fake images
                    ######################################################
                    noise.data.normal_(0, 1)
                    fake_imgs, _, _, _ = netG(noise, sent_emb, words_embs, mask)
                    for j in range(batch_size):
                        s_tmp = '%s/single/%s' % (save_dir, keys[j])
                        folder = s_tmp[:s_tmp.rfind('/')]
                        if not os.path.isdir(folder):
                            print('Make a new folder: ', folder)
                            mkdir_p(folder)
                        k = -1
                        # for k in range(len(fake_imgs)):
                        im = fake_imgs[k][j].data.cpu().numpy()
                        # [-1, 1] --> [0, 255]
                        im = (im + 1.0) * 127.5
                        im = im.astype(np.uint8)
                        im = np.transpose(im, (1, 2, 0))
                        im = Image.fromarray(im)
                        fullpath = '%s_s%d.png' % (s_tmp, k)
                        im.save(fullpath)
      
                        
    def gen_example(self, data_dic):
        captions, cap_lens, sorted_indices = data_dic

        batch_size = captions.shape[0]
        nz = cfg.GAN.Z_DIM
        captions = Variable(torch.from_numpy(captions), volatile=True)
        cap_lens = Variable(torch.from_numpy(cap_lens), volatile=True)

        captions = captions.cuda()
        cap_lens = cap_lens.cuda()
           
        noise = Variable(torch.FloatTensor(batch_size, nz), volatile=True)
        noise = noise.cuda()
        #######################################################
        # (1) Extract text embeddings
        ######################################################
        hidden = self.text_encoder.init_hidden(batch_size)
        # words_embs: batch_size x nef x seq_len
        # sent_emb: batch_size x nef
        words_embs, sent_emb = self.text_encoder(captions, cap_lens, hidden)
        mask = (captions == 0)
        #######################################################
        # (2) Generate fake images
        ######################################################
        noise.data.normal_(0, 1)
        fake_imgs, attention_maps, _, _ = self.netG(noise, sent_emb, words_embs, mask)
        # G attention
        cap_lens_np = cap_lens.cpu().data.numpy()
        generated_images = []
        for j in range(batch_size):
            im = fake_imgs[2][j].data.cpu().numpy()
            im = (im + 1.0) * 127.5
            im = im.astype(np.uint8)
            # print('im', im.shape)
            im = np.transpose(im, (1, 2, 0))
            # print('im', im.shape)
            #im = Image.fromarray(im)
            #fullpath = '%s.png' % (save_name)
            #im.save(fullpath)
            generated_images.append(im)
           
        return np.array(generated_images)