parser.add_argument("--resume", action="store_true") opts = parser.parse_args() cudnn.benchmark = True # Load experiment setting config = get_config(opts.config) max_iter = config['max_iter'] if config['trainer'] == 'LipTrainer': trainer = LipTrainer(config) else: sys.exit('Train option not supported') trainer.cuda() train_loader = get_data_loader_list(config, train=True) model_name = config['trainer'] train_writer = tensorboardX.SummaryWriter( os.path.join(opts.output_path + "/logs", model_name)) output_directory = os.path.join(opts.output_path + "/outputs", model_name) checkpoint_directory = prepare_sub_folder(output_directory) shutil.copy(opts.config, os.path.join(output_directory, 'config.yaml')) iterations = trainer.resume(checkpoint_directory, param=config) if opts.resume else 0 while True: for id, data in enumerate(train_loader): trainer.update_learning_rate() audio = data['AU'].cuda(async=True).detach() parameter = data['PM'].cuda(async=True).detach()
trainer = GanimationTrainer(config) trainer.cuda() # Load audio2landmark module state_dict_lstm = torch.load(opts.checkpoint_lstm) trainer.audio2exp.load_state_dict(state_dict_lstm['audio2exp']) # Load neutral2emotion module state_dict_gan = torch.load(opts.checkpoint_n2e) trainer.gan.load_state_dict(state_dict_gan['gan']) # Load refinement module state_dict_gan = torch.load(opts.checkpoint_ref) trainer.encdec.load_state_dict(state_dict_gan['gan']) trainer.eval() test_loader = get_data_loader_list(config, train=False, demo=True) model_name = config['trainer'] train_writer = tensorboardX.SummaryWriter( os.path.join(opts.output_path + "/logs", model_name)) output_directory = os.path.join(opts.output_path + "/outputs", model_name) checkpoint_directory, image_directory = prepare_sub_folder(output_directory) with torch.no_grad(): for id, data in enumerate(test_loader): video_ref = data['VR'].cuda().detach() audio = data['AU'].cuda().detach() em = data['EM'][0] level = data['LV'][0] # Main training code tkfc = trainer.forward(video_ref, audio, em, level) #[0].cpu()
output_directory = os.path.join(opts.output_path + "/outputs", model_name + '_' + opts.tag) checkpoint_directory, image_directory, landmark_directory = prepare_sub_folder( output_directory, is_test=True) landmark_directory = 'datasets/face/test_keypoints/keypoints' shutil.copy(opts.config, os.path.join(output_directory, 'config.yaml')) f = open(os.path.join(root_dir, config['pca_path']), 'rb') pca = pickle.load(f) trainer = LipTrainer(config, is_train=False) trainer.to(config['device']) state_dict_lstm = torch.load(opts.checkpoint_lstm) trainer.audio2exp.load_state_dict(state_dict_lstm['audio2exp']) test_loader = get_data_loader_list(config, split='test') iterations = trainer.resume(checkpoint_directory, param=config) if opts.resume else 0 loss_test = 0 cnt = 0 for id, data in enumerate(test_loader): audio = data[0].to(config['device']).detach() target_kp = data[1].to(config['device']).detach() items = len(audio) N = data[2][0] # (-, N, theta, mean, ...) theta = data[3][0] mean = data[4][0].numpy() all_ldmk = data[5][0].numpy() frame_id = data[6][0] img = data[7][0].numpy()
type=str, default='.', help="outputs path") parser.add_argument("--resume", action="store_true") opts = parser.parse_args() cudnn.benchmark = True # Load experiment setting config = get_config(opts.config) max_iter = config['max_iter'] trainer = GanimationTrainer(config) trainer.cuda() train_loader = get_data_loader_list(config, train=True, demo=False) model_name = config['trainer'] train_writer = tensorboardX.SummaryWriter( os.path.join(opts.output_path + "/logs", model_name)) output_directory = os.path.join(opts.output_path + "/outputs", model_name) checkpoint_directory, image_directory = prepare_sub_folder(output_directory) shutil.copy(opts.config, os.path.join(output_directory, 'config.yaml')) iterations = trainer.resume(checkpoint_directory, param=config) if opts.resume else 0 while True: for id, data in enumerate(train_loader): trainer.update_learning_rate() # label data emc_label = data['EL'].cuda().detach() int_label = data['IL'].cuda().detach()
default='a2l', help="outputs path") parser.add_argument("--resume", action="store_true") opts = parser.parse_args() cudnn.benchmark = True # Load experiment setting config = get_config(opts.config) max_iter = config['max_iter'] trainer = LipTrainer(config) trainer.to(config['device']) train_loader = get_data_loader_list(config, split='train') eval_loader = get_data_loader_list(config, split='eval') model_name = config['trainer'] # train_writer = tensorboardX.SummaryWriter(os.path.join(opts.output_path + "/logs", model_name)) logging.basicConfig(filename=os.path.join(opts.output_path + "/logs", model_name + '.log'), level=logging.DEBUG) logging.debug('This message should go to the log file') logging.info('So should this') output_directory = os.path.join(opts.output_path + "/outputs", model_name) checkpoint_directory = os.path.join(prepare_sub_folder(output_directory), '..') shutil.copy(opts.config, os.path.join(output_directory, 'config.yaml')) iterations = trainer.resume(checkpoint_directory, param=config) if opts.resume else 0