- img = tf.placeholder(name="img", dtype=tf.float32, shape=(1, feature_map_size, feature_map_size, input_channels))
- util.generate_model(self.get_model_path(), normalize_keras_wrapper(tf.keras.layers.Conv2D), img, filters=output_channels, kernel_size=(ksize, ksize), strides=stride, padding='same', name='the_conv')
复制代码
- >>> inputs
- [<tf.Operation 'img' type=Placeholder>]
- >>> inputs[0].outputs
- [<tf.Tensor 'img:0' shape=(1, 7, 7, 1916) dtype=float32>]
- >>> outputs
- [<tf.Operation 'the_conv/BiasAdd' type=BiasAdd>]
- >>> outputs[0].outputs
- [<tf.Tensor 'the_conv/BiasAdd:0' shape=(1, 7, 7, 320) dtype=float32>]
复制代码
欢迎光临 Toybrick (https://t.rock-chips.com/) | Powered by Discuz! X3.3 |