little changes

  def __init__(self, n_tasks, n_features, logdir=None, layer_sizes=[1000],

               weight_init_stddevs=[.02], bias_init_consts=[1.], penalty=0.0,

               penalty_type="l2", dropouts=[0.5], learning_rate=.001,

	       learning_rate_decay = 0.9, learning_rate_decay_start = 5,

               momentum=".9", optimizer="adam", batch_size=50, n_classes=2,

               train=True, verbosity=None, **kwargs):

    """Constructs the computational graph.

    self.penalty_type = penalty_type

    self.dropouts = dropouts

    self.learning_rate = learning_rate

    self.learning_rate_decay = learning_rate_decay

    self.learning_rate_decay_start = learning_rate_decay_start 

    self.momentum = momentum

    self.optimizer = optimizer

    self.batch_size = batch_size

        # Save an initial checkpoint.

        saver.save(sess, self._save_path, global_step=0)

        for epoch in range(nb_epoch):

          avg_loss, n_batches = 0., 0

          if shuffle:

            log("About to shuffle dataset before epoch start.", self.verbosity)

    #running model

    for i, hp in enumerate(hyper_parameters[model]):

    for count, hp in enumerate(hyper_parameters[model]):

      time_start_fitting = time.time()

      train_score,valid_score = benchmark_train_and_valid(base_dir,

                                    train_dataset, valid_dataset, tasks,

      time_finish_fitting = time.time()

      with open(os.path.join(out_path,'results.csv'),'a') as f:

        f.write('\n\n'+str(i))

        f.write('\n\n'+str(count))

        f.write('\n'+dname+',train')

        for i in train_score:

          f.write(','+i+','+str(train_score[i]['mean-roc_auc_score']))

  hps['rf'] = [{'n_estimators':500}]

  benchmark_loading_datasets(base_dir_o,hps,n_features = 1024,

                             dataset_name = dataset_name, model = model,

                             reload = reload, verbosity = 'high')

  benchmark_loading_datasets(base_dir_o,hps,dataset_name = dataset_name,

                             model = model,reload = reload,verbosity = 'high')