Hello, when i train the model with my own dataset, i found the statement:
dropout = tf.layers.dropout(pool, rate=drop_rate)
in model.py may be have problem.
The operation: tf.layers.dropout() in TensorFlow's document has an argument of training, and training=False by default, in your code, you didn't set the training argument, so, although you set the rate value, the model won't conduct the dropout operation???
Hello, when i train the model with my own dataset, i found the statement:
dropout = tf.layers.dropout(pool, rate=drop_rate)in model.py may be have problem.
The operation: tf.layers.dropout() in TensorFlow's document has an argument of training, and training=False by default, in your code, you didn't set the training argument, so, although you set the rate value, the model won't conduct the dropout operation???