tensorboard添加在程序中出现的错误

tensorboard 在程序中出现的错误：

1.InvalidArgumentError: Shape [-1,2352] has negative dimensions
2.FailedPreconditionError: Attempting to use uninitialized value local3/weight

错误代码段1：

#合并 summary
summary_op = tf.summary.merge_all()
sess = tf.Session()
train_writer = tf.summary.FileWriter(logs_train_dir, sess.graph)
with tf.Session() as sess：
     sess.run(tf.global_variables_initializer())
      ..........省略...........
      summary_str = sess.run(summary_op)
      train_writer.add_summary(summary_str, iteration)

出现错误1：InvalidArgumentError: Shape [-1,2352] has negative dimensions

错误代码段2：

# 合并 summary
summary_op = tf.summary.merge_all()
sess = tf.Session()
# 保存summary
train_writer = tf.summary.FileWriter(logs_train_dir, sess.graph)
sess.run(tf.global_variables_initializer())
with tf.Session() as sess:
      ..........省略...........
      summary_str = sess.run(summary_op)
      train_writer.add_summary(summary_str, iteration)
      

出现错误2：FailedPreconditionError: Attempting to use uninitialized value local3/weight

解决办法：

    sess = tf.Session()
    # 合并 summary
    summary_op = tf.summary.merge_all()
    # 保存summary
    train_writer = tf.summary.FileWriter(logs_train_dir, sess.graph)
    sess.run(tf.global_variables_initializer()) 
 ........省略.........
            summary_str = sess.run(summary_op,feed_dict = feed)
            train_writer.add_summary(summary_str, iteration)

将tf.Session 提到tf.summary.merge_all()的前面
在sess.run(summary_op)时喂入feed_dict，不喂入会出现未对张量赋值。
参考文献：莫烦GitHub_tensorboard_test