fwu11的博客

导入预训练模型进行推理。以mobilenetv2 Imagenet预训练模型为例：import tensorflow as tfimport tensorflow_hub as hubimport numpy as np# 加载图片image = tf.image.decode_image(tf.io.read_file('/path/to/your/image'),channels=3)image = tf.image.resize(image,(224,224))image = tf

以mobilenet v2为例import tensorflow as tfimport tensorflow_hub as hubIMAGE_SHAPE = (224, 224)m = tf.keras.Sequential([ hub.KerasLayer("https://hub.tensorflow.google.cn/google/tf2-preview/mobilenet_v2/classification/4", input_shape=IMAGE_SHAPE+(3,),

想把saved_model.pb里的metagraph转成可读。代码如下：#将.pb文件load到session中，导出到.pbtxt可视化import tensorflow as tffrom tensorflow.python.platform import gfiletf.compat.v1.disable_eager_execution()model = './savedmodel/VGG16/1/'export_dir = './test/'with tf.compat.v1.