diff options
| -rwxr-xr-x | models/pix2pixHD_model.py | 16 | ||||
| -rwxr-xr-x | train.py | 15 |
2 files changed, 22 insertions, 9 deletions
diff --git a/models/pix2pixHD_model.py b/models/pix2pixHD_model.py index 631a10f..79ebabd 100755 --- a/models/pix2pixHD_model.py +++ b/models/pix2pixHD_model.py @@ -11,7 +11,13 @@ from . import networks class Pix2PixHDModel(BaseModel): def name(self): return 'Pix2PixHDModel' - + + def init_loss_filter(self, use_gan_feat_loss, use_vgg_loss): + flags = (True, use_gan_feat_loss, use_vgg_loss, True, True) + def loss_filter(g_gan, g_gan_feat, g_vgg, d_real, d_fake): + return [l for (l,f) in zip((g_gan,g_gan_feat,g_vgg,d_real,d_fake),flags) if f] + return loss_filter + def initialize(self, opt): BaseModel.initialize(self, opt) if opt.resize_or_crop != 'none': # when training at full res this causes OOM @@ -20,7 +26,6 @@ class Pix2PixHDModel(BaseModel): self.use_features = opt.instance_feat or opt.label_feat self.gen_features = self.use_features and not self.opt.load_features input_nc = opt.label_nc if opt.label_nc != 0 else 3 - ##### define networks # Generator network netG_input_nc = input_nc @@ -65,13 +70,16 @@ class Pix2PixHDModel(BaseModel): self.old_lr = opt.lr # define loss functions + self.loss_filter = self.init_loss_filter(not opt.no_ganFeat_loss, not opt.no_vgg_loss) + self.criterionGAN = networks.GANLoss(use_lsgan=not opt.no_lsgan, tensor=self.Tensor) self.criterionFeat = torch.nn.L1Loss() if not opt.no_vgg_loss: self.criterionVGG = networks.VGGLoss(self.gpu_ids) + # Names so we can breakout loss - self.loss_names = ['G_GAN', 'G_GAN_Feat', 'G_VGG', 'D_real', 'D_fake'] + self.loss_names = self.loss_filter('G_GAN','G_GAN_Feat','G_VGG','D_real', 'D_fake') # initialize optimizers # optimizer G @@ -175,7 +183,7 @@ class Pix2PixHDModel(BaseModel): loss_G_VGG = self.criterionVGG(fake_image, real_image) * self.opt.lambda_feat # Only return the fake_B image if necessary to save BW - return [ [ loss_G_GAN, loss_G_GAN_Feat, loss_G_VGG, loss_D_real, loss_D_fake ], None if not infer else fake_image ] + return [ self.loss_filter( loss_G_GAN, loss_G_GAN_Feat, loss_G_VGG, loss_D_real, loss_D_fake ), None if not infer else fake_image ] def inference(self, label, inst): # Encode Inputs @@ -38,7 +38,12 @@ print('#training images = %d' % dataset_size) model = create_model(opt) visualizer = Visualizer(opt) -total_steps = (start_epoch-1) * dataset_size + epoch_iter +total_steps = (start_epoch-1) * dataset_size + epoch_iter + +display_delta = total_steps % opt.display_freq +print_delta = total_steps % opt.print_freq +save_delta = total_steps % opt.save_latest_freq + for epoch in range(start_epoch, opt.niter + opt.niter_decay + 1): epoch_start_time = time.time() if epoch != start_epoch: @@ -49,7 +54,7 @@ for epoch in range(start_epoch, opt.niter + opt.niter_decay + 1): epoch_iter += opt.batchSize # whether to collect output images - save_fake = total_steps % opt.display_freq == 0 + save_fake = total_steps % opt.display_freq == display_delta ############## Forward Pass ###################### losses, generated = model(Variable(data['label']), Variable(data['inst']), @@ -61,7 +66,7 @@ for epoch in range(start_epoch, opt.niter + opt.niter_decay + 1): # calculate final loss scalar loss_D = (loss_dict['D_fake'] + loss_dict['D_real']) * 0.5 - loss_G = loss_dict['G_GAN'] + loss_dict['G_GAN_Feat'] + loss_dict['G_VGG'] + loss_G = loss_dict['G_GAN'] + loss_dict.get('G_GAN_Feat',0) + loss_dict.get('G_VGG',0) ############### Backward Pass #################### # update generator weights @@ -78,7 +83,7 @@ for epoch in range(start_epoch, opt.niter + opt.niter_decay + 1): ############## Display results and errors ########## ### print out errors - if total_steps % opt.print_freq == 0: + if total_steps % opt.print_freq == print_delta: errors = {k: v.data[0] if not isinstance(v, int) else v for k, v in loss_dict.items()} t = (time.time() - iter_start_time) / opt.batchSize visualizer.print_current_errors(epoch, epoch_iter, errors, t) @@ -92,7 +97,7 @@ for epoch in range(start_epoch, opt.niter + opt.niter_decay + 1): visualizer.display_current_results(visuals, epoch, total_steps) ### save latest model - if total_steps % opt.save_latest_freq == 0: + if total_steps % opt.save_latest_freq == save_delta: print('saving the latest model (epoch %d, total_steps %d)' % (epoch, total_steps)) model.module.save('latest') np.savetxt(iter_path, (epoch, epoch_iter), delimiter=',', fmt='%d') |