D0 = model.E_Net_2() encoder_model = torch.load("./Dmodel") # 24*24 enet = model.E_Net_2().cuda() enet.load_state_dict(encoder_model) add_in_feature = 240+ hidden_d # Add one dimension data for the input_feature data. gnet = model.G_Net_FM_3(ngpu,add_in_feature,main_gpu=main_gpu).cuda() # g_model = torch.load("./fm21/G_95000.model") # gnet.load_state_dict(g_model) d_in_demension = 2 dnet = model.D_Net_conv(ngpu,d_in_demension,main_gpu=main_gpu).cuda() # nosie_d = 10 g_net_optimizer = optim.Adam(gnet.parameters(),lr = 1e-4) d_net_optimizer = optim.Adam(dnet.parameters(), lr = 1e-4) check_points = 500 num_epoches = 100000 criterion = nn.BCELoss() for epoch in (range(1,num_epoches)): # print("give me a clue") # data, label = mnist.train.next_batch(batch_size) data, label = mm.batch_next(batch_size, [0, 1, 2, 3, 4, 5, 6, 7, 8, 9], shuffle=True) dnet.zero_grad() gnet.zero_grad()
mm = data_convert.owndata() num = '0' out_dir = 'out_fc_{}_{}/'.format(datetime.now(), num) if not os.path.exists(out_dir): os.makedirs(out_dir) shutil.copyfile(sys.argv[0], out_dir + '/training_script.py') # else: # print("you have already creat one.") # exit(1) sys.stdout = mutil.Logger(out_dir) # in_channel = 1 G = model.G_Net_conv(ngpu).cuda() D = model.D_Net_conv(ngpu, in_channel).cuda() """Weight Initialization""" def weights_init(m): classname = m.__class__.__name__ if classname.find('Conv') != -1: m.weight.data.normal_(0.0, 0.02) # elif classname.find('BatchNorm') != -1: # m.weight.data.normal_(1.0, 0.02) # m.bias.data.fill_(0) G.apply(weights_init) # G.load_state_dict(torch.load('./out_conv_part/G_20000.model')) D.apply(weights_init)
out_dir = './cifar100_result/basic_{}_{}/'.format(datetime.now(), num) out_dir.replace(" ", "_") if not os.path.exists(out_dir): os.makedirs(out_dir) shutil.copyfile(sys.argv[0], out_dir + '/training_script.py') sys.stdout = mutil.Logger(out_dir) in_channel = 4 d_num = 3 G = model.G_Net_conv_32(ngpu, main_gpu=gpu, in_channel=Z_dim + label_dim, out_channel=3).cuda() D = model.D_Net_conv(ngpu, in_channel, main_gpu=gpu).cuda() """Weight Initialization""" # def weights_init(m): # classname = m.__class__.__name__ # if classname.find('Conv') != -1: # m.weight.data.normal_(0.0, 0.02) """ ===================== TRAINING ======================== """ d_num = 3 # avd_num = 1/d_num G_solver = optim.Adam(G.parameters(), lr=1e-4) D_solver = optim.Adam(D.parameters(), lr=2e-5) ones_label = Variable(torch.ones(mb_size)).cuda() zeros_label = Variable(torch.zeros(mb_size)).cuda()
c_label[i:(i + 6), i / 6] = 1. sys.stdout = mutil.Logger(out_dir) # else: # print("you have already creat one.") # exit(1) # # # def xavier_init(size): # in_dim = size[0] # xavier_stddev = 1. / np.sqrt(in_dim / 2.) # return Variable(torch.randn(*size) * xavier_stddev, requires_grad=True) in_channel = 2 G = model.G_Net_conv(ngpu).cuda() D = model.D_Net_conv(ngpu, 1).cuda() E = model.Ev_Net_conv(ngpu, 1).cuda() """Weight Initialization""" def weights_init(m): classname = m.__class__.__name__ if classname.find('Conv') != -1: m.weight.data.normal_(0.0, 0.02) # elif classname.find('BatchNorm') != -1: # m.weight.data.normal_(1.0, 0.02) # m.bias.data.fill_(0) G.apply(weights_init) D.apply(weights_init)
cnt = 0 num = '0' out_dir = './cifar100_result/basic_{}_{}/'.format(datetime.now(), num) out_dir.replace(" ", "_") if not os.path.exists(out_dir): os.makedirs(out_dir) shutil.copyfile(sys.argv[0], out_dir + '/training_script.py') sys.stdout = mutil.Logger(out_dir) in_channel = 2 d_num = 3 G = model.G_Net_conv(ngpu).cuda() D_list = [model.D_Net_conv(ngpu, in_channel).cuda() for i in range(d_num)] """Weight Initialization""" def weights_init(m): classname = m.__class__.__name__ if classname.find('Conv') != -1: m.weight.data.normal_(0.0, 0.02) """ ===================== TRAINING ======================== """ d_num = 3 # avd_num = 1/d_num G_solver = optim.Adam(G.parameters(), lr=1e-4)