利用TensorFlow進行模型finetune時，想訓練指定的op，則須要根據指定的op名進行提取，而後將其配置至tf.train.AdamOptimizer函數中的var_list參數。python
部分代碼以下：函數
def _get_variables_to_train(trainable_scopes = None):
"""Returns a list of variables to train.
Returns:
A list of variables to train by the optimizer.
"""
if trainable_scopes is None:
return tf.trainable_variables()
else:
scopes = [scope.strip() for scope in trainable_scopes.split(',')]
variables_to_train = []
for scope in scopes:
variables = tf.get_collection(tf.GraphKeys.TRAINABLE_VARIABLES, scope)
variables_to_train.extend(variables)
return variables_to_train
output_vars = _get_variables_to_train(Config.trainable_scopes)
update_ops = tf.get_collection(tf.GraphKeys.UPDATE_OPS)
with tf.control_dependencies(update_ops):
train_op = tf.train.AdamOptimizer(learning_rate=lr).minimize(loss, global_step=global_step, var_list=output_vars)
trainable_scopes = 'HCnet/Bottle_neck5,HCnet/Bottle_neck5_1,HCnet/Bottle_neck6_2,HCnet/Conv7'
trainable_scopes是我須要訓練的op，經過函數_get_variables_to_train獲取須要的op參數，而後將獲取到的參數output_vars傳入tf.train.AdamOptimizer中的var_list中，經過該方法便可對指定的op訓練。上述過程省略了模型預加載的過程。spa