def __init__(self, net, hidden_dim, num_gc_layers, alpha=0.5, beta=1., gamma=.1, projection_size = 256, projection_hidden_size = 4096, moving_average_decay = 0.99): super(CLSA, self).__init__() self.alpha = alpha self.beta = beta self.gamma = gamma args = arg_parse() self.prior = args.prior self.embedding_dim = hidden_dim * num_gc_layers self.online_encoder = net self.target_encoder = None self.target_ema_updater = EMA(moving_average_decay) self.online_predictor = MLP(projection_size, projection_size, projection_hidden_size) self.init_emb()
def seed_everything(seed=1234): random.seed(seed) torch.manual_seed(seed) torch.cuda.manual_seed_all(seed) np.random.seed(seed) os.environ['PYTHONHASHSEED'] = str(seed) torch.backends.cudnn.deterministic = True torch.backends.cudnn.benchmark = False if __name__ == '__main__': seed_everything() from model import Net from arguments import arg_parse args = arg_parse() target = args.target dim = 64 epochs = 500 batch_size = 20 lamda = args.lamda use_unsup_loss = args.use_unsup_loss separate_encoder = args.separate_encoder path = osp.join(osp.dirname(osp.realpath(__file__)), '..', 'data', 'QM9') transform = T.Compose([MyTransform(), Complete(), T.Distance(norm=False)]) dataset = QM9(path, transform=transform).shuffle() print('num_features : {}\n'.format(dataset.num_features)) # Normalize targets to mean = 0 and std = 1.