Monitors are deprecated. Please use tf.train.SessionRunHook.

[ghost]

while I run the udc_train.py on tensorflow 1.1.0, it raise the error like that:

WARNING:tensorflow:From C:\Anaconda3\lib\site-packages\tensorflow\contrib\learn\python\learn\monitors.py:267: BaseMonitor.init (from tensorflow.contrib.learn.python.learn.monitors) is deprecated and will be removed after 2016-12-05. Instructions for updating: Monitors are deprecated. Please use tf.train.SessionRunHook. Traceback (most recent call last): File "E:/project/chatbot/udc_train.py", line 67, in tf.app.run() File "C:\Anaconda3\lib\site-packages\tensorflow\python\platform\app.py", line 48, in run _sys.exit(main(_sys.argv[:1] + flags_passthrough)) File "E:/project/chatbot/udc_train.py", line 64, in main estimator.fit(input_fn=input_fn_train, steps=None, monitors=[eval_monitor]) File "C:\Anaconda3\lib\site-packages\tensorflow\python\util\deprecation.py", line 281, in new_func return func(*args, **kwargs) File "C:\Anaconda3\lib\site-packages\tensorflow\contrib\learn\python\learn\estimators\estimator.py", line 430, in fit loss = self._train_model(input_fn=input_fn, hooks=hooks) File "C:\Anaconda3\lib\site-packages\tensorflow\contrib\learn\python\learn\estimators\estimator.py", line 927, in _train_model model_fn_ops = self._get_train_ops(features, labels) File "C:\Anaconda3\lib\site-packages\tensorflow\contrib\learn\python\learn\estimators\estimator.py", line 1132, in _get_train_ops return self._call_model_fn(features, labels, model_fn_lib.ModeKeys.TRAIN) File "C:\Anaconda3\lib\site-packages\tensorflow\contrib\learn\python\learn\estimators\estimator.py", line 1103, in _call_model_fn model_fn_results = self._model_fn(features, labels, **kwargs) File "E:\project\chatbot\udc_model.py", line 39, in model_fn targets File "E:\project\chatbot\models\dual_encoder.py", line 47, in dual_encoder_model tf.concat(0, [context_embedded, utterance_embedded]), File "C:\Anaconda3\lib\site-packages\tensorflow\python\ops\array_ops.py", line 1029, in concat dtype=dtypes.int32).get_shape( File "C:\Anaconda3\lib\site-packages\tensorflow\python\framework\ops.py", line 639, in convert_to_tensor as_ref=False) File "C:\Anaconda3\lib\site-packages\tensorflow\python\framework\ops.py", line 704, in internal_convert_to_tensor ret = conversion_func(value, dtype=dtype, name=name, as_ref=as_ref) File "C:\Anaconda3\lib\site-packages\tensorflow\python\framework\constant_op.py", line 113, in _constant_tensor_conversion_function return constant(v, dtype=dtype, name=name) File "C:\Anaconda3\lib\site-packages\tensorflow\python\framework\constant_op.py", line 102, in constant tensor_util.make_tensor_proto(value, dtype=dtype, shape=shape, verify_shape=verify_shape)) File "C:\Anaconda3\lib\site-packages\tensorflow\python\framework\tensor_util.py", line 370, in make_tensor_proto _AssertCompatible(values, dtype) File "C:\Anaconda3\lib\site-packages\tensorflow\python\framework\tensor_util.py", line 302, in _AssertCompatible (dtype.name, repr(mismatch), type(mismatch).name)) TypeError: Expected int32, got list containing Tensors of type '_Message' instead.

could you give me some advice, thank you

[rmandyam]

The order of arguments for tf.concat and tf.split have changed in TF > 0.12. So, the actual error is in line 47 dual_encoder_model.. Replace these lines below ( # commented) and corrected below. This should help.. Also, you may have to replace tf.batch_matmul with tf.matmul and losses = tf.nn.sigmoid_cross_entropy_with_logits(logits=logits, labels=tf.to_float(targets))

rnn_outputs, rnn_states = tf.nn.dynamic_rnn( cell, #tf.concat(0, [context_embedded, utterance_embedded]), tf.concat( [context_embedded, utterance_embedded], 0), #sequence_length=tf.concat(0, [context_len, utterance_len]), sequence_length=tf.concat([context_len, utterance_len], 0), dtype=tf.float32) #encoding_context, encoding_utterance = tf.split(0, 2, rnn_states.h) encoding_context, encoding_utterance = tf.split(rnn_states.h, 2 , 0)

[jayurbain]

Fixing tf.concat does not fix this problem.

[rmandyam]

No, it fixes the problem in the trace above at line 47 in dual_encoder.py, which is due to a change in the order of arguments for tf.concat and tf.split in versions of TF > 0.12 There is a disconnect between the title of this issue and the actual stack trace.
The solution for Monitors being deprecated is to use tf.train.SessionRunHook. It's in the title itself.

[nikky78]

How do you use tf.train.SessionRunHook ? for instance I have : hooks = [tf.train.LoggingTensorHook({'loss'}, every_n_iter = 4)]

learn.Experiment(estimator = estimator, train_input_fn = get_train_inputs, eval_input_fn = get_eval_inputs, train_monitors = hooks, eval_hooks = hooks)

It says: Monitors are deprecated. Please use tf.train.SessionRunHook., but I don't know how to change it.

[logust79]

I have same problem as @nikky78

[RahulRishavMohanti]

You don't really have to @nikky78 , @logust79 . All you gotta do is change the order of concat and split parameters as mentioned before, replace batch_matmul with matmul and then alter line 81 in dual_encoder.py losses = tf.nn.sigmoid_cross_entropy_with_logits(logits, tf.to_float(targets)) with losses = tf.nn.sigmoid_cross_entropy_with_logits(logits=logits, labels=tf.to_float(targets))