tensorflow pretrained model

tensorflow 简单的加载训练好的模型，如果某一层形状不一样就自动的省略，也可以用reshape。

def optimistic_restore(session, save_file):
    reader = tf.train.NewCheckpointReader(save_file)
    saved_shapes = reader.get_variable_to_shape_map()
    var_names = sorted([(var.name, var.name.split(':')[0]) for var in tf.global_variables()
            if var.name.split(':')[0] in saved_shapes])
    restore_vars = []
    with tf.variable_scope('', reuse=True):
        for var_name, saved_var_name in var_names:
            curr_var = tf.get_variable(saved_var_name)
            var_shape = curr_var.get_shape().as_list()
            if var_shape == saved_shapes[saved_var_name]:
                restore_vars.append(curr_var)
    saver = tf.train.Saver(restore_vars)
    saver.restore(session, save_file)

reference
https://github.com/tensorflow/tensorflow/issues/312