import tensorflow as tf
import sys
 
trained_checkpoint_prefix = sys.argv[1]
export_dir = sys.argv[2]
graph = tf.Graph()
config=tf.ConfigProto(allow_soft_placement=True, log_device_placement=True)
with tf.compat.v1.Session(graph=graph, config=config) as sess:
  # Restore from checkpoint
  loader = tf.compat.v1.train.import_meta_graph(trained_checkpoint_prefix + '.meta')
  loader.restore(sess, trained_checkpoint_prefix)
 
  # Export checkpoint to SavedModel
  builder = tf.compat.v1.saved_model.builder.SavedModelBuilder(export_dir)
  builder.add_meta_graph_and_variables(sess, [tf.saved_model.TRAINING, tf.saved_model.SERVING], strip_default_attrs=True)
  builder.save()

from nets.inception_v1 import inception_v1, inception_v1_arg_scope

with slim.arg_scope(inception_v1_arg_scope()):
  logits, endpoints = inception_v1(
     images,
     dropout_keep_prob=0.8,
     num_classes=num_classes,
     is_training=for_training,
     scope=scope)

export CUDA_VISIBLE_DEVICES='0,1'  #shell环境
 
import os
os.environ["CUDA_VISIBLE_DEVICES"] = "0,1" #python环境