https://github.com/kongxindezhulhj/cnn-text-classification-master
参考:https://blog.csdn.net/edwards_june/article/details/65652385
需要修改兼容:
1.Python版本的问题:
FLAGS = tf.flags.FLAGS
FLAGS._parse_flags()
print("\nParameters:")
for attr, value in sorted(FLAGS.__flags.items()):
print("{}={}".format(attr.upper(), value))
print("")
改成:
FLAGS = tf.flags.FLAGS
#FLAGS._parse_flags()
print("\nParameters:")
for attr, value in sorted(FLAGS.flag_values_dict().items()):
print("{}={}".format(attr.upper(), value))
print("")
另外:
FLAGS.file_dir改成:
FLAGS.flag_values_dict()['file_dir']
2.tensorflow版本改动api bug报错:ValueError: Shapes (2, 100, 1) and () are incompatible
concated = tf.concat(1, [indices, sparse_labels])改为:
concated= tf.concat([indices, sparse_labels], 1)
3报错:
调用TensorFlow框架中的softmax_cross_entropy_with_logits()方法时,报了如下的错误:ValueError: Only call `softmax_cross_entropy_with_logits` with named arguments (labels=..., logits=..., ...)
cost_sme = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits, labels))改成:
cost_sme = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=logits, labels=labels))
3.报错:AttributeError: 'module' object has no attribute 'scalar_summary'
tf.scalar_summary('images', images)改为:tf.summary.scalar('images', images)tf.image_summary('images', images)改为:tf.summary.image('images', images)类似的有:tf.train.SummaryWriter改为:tf.summary.FileWritertf.merge_all_summaries()改为:summary_op = tf.summary.merge_alltf.histogram_summary(var.op.name, var)改为: tf.summary.histogramconcated = tf.concat(1, [indices, sparse_labels])改为:concated = tf.concat([indices, sparse_labels], 1)