Ranger-tensorflow Radam+lookahead implemented by tensorflow Usage from ranger import Ranger num_sample=1000 lr = tf.train.exponential_decay(1e-8, global_step, num_sample*10, 0.9, staircase=True) optimize = Ranger(learning_rate=lr,beta1=0.90,epsilon=1e-8)