def prior_update_gan_noise(particles, inputs, model_no):
netG = net.G(5, 3, 64)
netG.load_state_dict(
torch.load(
'%s/netG_epoch_%d.pth' %
('/home/gtx1080/Abduallah/pix2pix.pytorch/imglog/fluid_noise',
model_no)))
netG.eval()
netG.cuda()
mu = np.load(
'/home/gtx1080/Sync/Kun/30_min_ele/non_ele_whole/train/mu.npy')
var = np.load(
'/home/gtx1080/Sync/Kun/30_min_ele/non_ele_whole/train/var.npy')
particles = np.concatenate((particles, inputs), axis=1)
particles, label = normalization_test(particles, particles[:, 0:3, :, :],
mu, var)
particles = torch.tensor(particles, dtype=torch.float).cuda()
with torch.no_grad():
particles = netG(particles)
particles = np.array(particles.cpu())
particles[:,
0, :, :] = particles[:, 0, :, :] * np.sqrt(var[5, 0]) + mu[5, 0]
particles[:,
1, :, :] = particles[:, 1, :, :] * np.sqrt(var[6, 0]) + mu[6, 0]
particles[:,
2, :, :] = particles[:, 2, :, :] * np.sqrt(var[7, 0]) + mu[7, 0]
particles[:, 0, :, :] = np.maximum(particles[:, 0, :, :], 0)
return particles
mean=(0.5, 0.5, 0.5),
std=(0.5, 0.5, 0.5),
split='val',
shuffle=False,
seed=opt.manualSeed)
# get logger
trainLogger = open('%s/train.log' % opt.exp, 'w')
ngf = opt.ngf
ndf = opt.ndf
inputChannelSize = opt.inputChannelSize
outputChannelSize = opt.outputChannelSize
# get models
netG = net.G(inputChannelSize, outputChannelSize, ngf)
netG.apply(weights_init)
if opt.netG != '':
netG.load_state_dict(torch.load(opt.netG))
print(netG)
netD = net.D(inputChannelSize + outputChannelSize, ndf)
netD.apply(weights_init)
if opt.netD != '':
netD.load_state_dict(torch.load(opt.netD))
print(netD)
netG.train()
netD.train()
criterionBCE = nn.BCELoss()
criterionCAE = nn.L1Loss()
seed=opt.manualSeed)
ngf = opt.ngf
ndf = opt.ndf
inputChannelSize = opt.inputChannelSize
outputChannelSize = opt.outputChannelSize
# get models
#import pdb; pdb.set_trace()
netG_BEGAN = netBEGAN.G(inputChannelSize, outputChannelSize, ngf)
netG_BEGAN.apply(weights_init)
if opt.netG_BEGAN != '':
netG_BEGAN.load_state_dict(torch.load(opt.netG_BEGAN))
print(netG_BEGAN)
netG_GAN = netGAN.G(inputChannelSize, outputChannelSize, ngf)
netG_GAN.apply(weights_init)
if opt.netG_GAN != '':
netG_GAN.load_state_dict(torch.load(opt.netG_GAN))
print(netG_GAN)
netG_BEGAN.train()
netG_GAN.train()
val_target = torch.FloatTensor(opt.tstBatchSize, outputChannelSize,
opt.imageSize, opt.imageSize)
val_input = torch.FloatTensor(opt.tstBatchSize, inputChannelSize,
opt.imageSize, opt.imageSize)
netG_BEGAN.cuda()
netG_GAN.cuda()
