def checkpoint_session(model, input_tensor_names, output_tensor_names,
**kwargs):
assert output_tensor_names is not None and len(output_tensor_names) > 0, \
'outputs should not be None of checkpoint....'
ckpt_prefix = [os.path.splitext(i)[0] for i in os.listdir(model) \
if i.endswith(".meta")][0]
config = tf.compat.v1.ConfigProto()
config.use_per_session_threads = 1
config.inter_op_parallelism_threads = 1
graph = tf.Graph()
sess = tf.compat.v1.Session(graph=graph, config=config)
with graph.as_default():
saver = tf.compat.v1.train.import_meta_graph(\
os.path.join(model, ckpt_prefix + '.meta'), clear_devices=True)
sess.run(tf.compat.v1.global_variables_initializer())
saver.restore(sess, os.path.join(model, ckpt_prefix))
from lpot.adaptor.tf_utils.util import get_input_node_names
input_tensor_names = input_tensor_names if validate_graph_node(\
sess.graph.as_graph_def(), tensor_to_node(input_tensor_names)) else \
get_input_node_names(sess.graph.as_graph_def())
return sess, input_tensor_names, output_tensor_names
def validate_and_inference_input_output(graph_def, \
input_tensor_names, output_tensor_names):
from lpot.adaptor.tf_utils.util import get_input_node_names
input_tensor_names = input_tensor_names if validate_graph_node(\
graph_def, tensor_to_node(input_tensor_names)) else \
get_input_node_names(graph_def)
from lpot.adaptor.tf_utils.util import get_output_node_names
output_tensor_names = output_tensor_names if validate_graph_node(\
graph_def, tensor_to_node(output_tensor_names)) else \
get_output_node_names(graph_def)
return input_tensor_names, output_tensor_names
def validate_and_inference_input_output(graph_def, \
input_tensor_names, output_tensor_names):
"""validate and inference the input and output tensor names of graph_def
Args:
graph_def (tf.compat.v1.GraphDef): tf.compat.v1.GraphDef object
input_tensor_names (list of string): input_tensor_names of graph_def
output_tensor_names (list of string): output_tensor_names of graph_def
Returns:
input_tensor_names (list of string): validated input_tensor_names
output_tensor_names (list of string): validated output_tensor_names
"""
from lpot.adaptor.tf_utils.util import get_input_node_names
input_tensor_names = input_tensor_names if validate_graph_node(\
graph_def, tensor_to_node(input_tensor_names)) else \
get_input_node_names(graph_def)
from lpot.adaptor.tf_utils.util import get_output_node_names
output_tensor_names = output_tensor_names if validate_graph_node(\
graph_def, tensor_to_node(output_tensor_names)) else \
get_output_node_names(graph_def)
return input_tensor_names, output_tensor_names