我想按如下方式运行张量流代码:
# Construct graph
g = Graph("train"); print("Graph loaded")
# Start session
sv = tf.train.Supervisor(graph=g.graph,logdir="/workspace",save_model_secs=0)
with sv.managed_session() as sess:
for epoch in range(1, 10):
if sv.should_stop(): break
for step in tqdm(range(g.num_batch), total=g.num_batch, ncols=70,
leave=False, unit='b'):
sess.run(g.train_op)
等待很长时间后,控制台会提示我错误:
ValueError:GraphDef不能大于2GB。 错误代码行是
sv=tf.train.Supervisor(graph=g.graph,logdir="/workspace",save_model_secs=0)
从控制台提示来看,似乎图形大小大于2GB。我该如何解决这个问题?