runtime_config.py

import tensorflow as tf

flags = tf.compat.v1.flags
# flags = tf.app.flags
FLAGS = flags.FLAGS
flags.DEFINE_string('model', 'gcn_cheby', 'Model string.')  # 'gcn', 'gcn_cheby', 'dense'
flags.DEFINE_float('learning_rate', 0.001, 'Initial learning rate.')
flags.DEFINE_float('learning_decay', 1.0, 'Initial learning rate.')
flags.DEFINE_integer('epochs', 201, 'Number of epochs to train.')
flags.DEFINE_integer('feature_size', 32, 'Number of units input feature.')
flags.DEFINE_integer('hidden1', 32, 'Number of units in hidden layer 1.')
flags.DEFINE_integer('diver_num', 32, 'Number of outputs.')
flags.DEFINE_float('dropout', 0, 'Dropout rate (1 - keep probaNUmbility).')
flags.DEFINE_float('weight_decay', 5e-4, 'Weight for L2 loss on embedding matrix.')
flags.DEFINE_integer('early_stopping', 1000, 'Tolerance for early stopping (# of epochs).')
flags.DEFINE_integer('max_degree', 1, 'Maximum Chebyshev polynomial degree.')
flags.DEFINE_integer('num_layer', 20, 'number of layers.')

flags.DEFINE_float('backoff_prob', 0.3, 'Probability of saving intermediate node on search tree.')
flags.DEFINE_integer('diver_out', 32, 'Number of predictors in test.')
flags.DEFINE_integer('timeout', 300, 'Total Seconds to Run each case.')
flags.DEFINE_string('datapath','./data/Random_Graph_Test', 'Location of Data')
flags.DEFINE_float('snr_db', 10, 'ratio of powers between weights and noise (dB).')
flags.DEFINE_string('training_set', 'IS4SAT', 'Name of training dataset')
flags.DEFINE_integer('greedy', 0, 'Normal: 0, Greedy: 1, Noisy Greedy: 2')
flags.DEFINE_bool('skip', False, 'If skip connection included')
flags.DEFINE_string('wts_init', 'random', 'how to initialize the weights of GCN')
flags.DEFINE_string('snapshot', '', 'snapshot of model')
flags.DEFINE_string('predict', 'mwis', 'direct output: mwis, linear combination: mis')

flags.DEFINE_float('epsilon', 1.0, 'initial exploration rate')
flags.DEFINE_float('epsilon_min', 0.001, 'minimal exploration rate')
flags.DEFINE_float('epsilon_decay', 0.985, 'exploration rate decay per replay')
flags.DEFINE_float('gamma', 1.0, 'gamma')