Closed zhengjxu closed 1 year ago
During the training, the keep_prob = 0.5, during inference, the keep_prob should be 1. I update the code by adding the keep_prob as tf.placeholder.
tf.placeholder
During the training, the keep_prob = 0.5, during inference, the keep_prob should be 1. I update the code by adding the keep_prob as
tf.placeholder
.