InvalidArgumentError (see above for traceback): TypeError: can't pickle dict_values objects

[programowalny]

System information

What is the top-level directory of the model you are using: object_detection
Have I written custom code (as opposed to using a stock example script provided in TensorFlow):
OS Platform and Distribution (e.g., Linux Ubuntu 16.04): Window 10
TensorFlow installed from (source or binary): conda installation
TensorFlow version (use command below): 1.8
Bazel version (if compiling from source): None
CUDA/cuDNN version: 9.0
GPU model and memory: GeForce GTX 1050TI (Laptop)
Exact command to reproduce: None

Describe the problem:

Fallow to instruction i want to train my images in 25 categories with e Faster-RCNN-Inception-V2-COCO after installation and configuration i typed

python model_main.py --logtostderr --train_dir=training/ --pipeline_config_path=training/faster_rcnn_inception_v2_pets.config
in my conda virtual enviroment, after a couple of minutes i get error presented above, i don't know, how to manage with this problem. In addition before a error i get warnning

WARNING:tensorflow:Ignoring ground truth with image id 1389737172 since it was previously added
WARNING:tensorflow:Ignoring detection with image id 1389737172 since it was previously added

But i don;t know if it is important.

Caused by op 'PyFunc_1', defined at:
  File "model_main.py", line 101, in <module>
    tf.app.run()
  File "C:\Users\damia\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\platform\app.py", line 125, in run
    _sys.exit(main(argv))
  File "model_main.py", line 97, in main
    tf.estimator.train_and_evaluate(estimator, train_spec, eval_specs[0])
  File "C:\Users\damia\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\estimator\training.py", line 447, in train_and_evaluate
    return executor.run()
  File "C:\Users\damia\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\estimator\training.py", line 531, in run
    return self.run_local()
  File "C:\Users\damia\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\estimator\training.py", line 681, in run_local
    eval_result, export_results = evaluator.evaluate_and_export()
  File "C:\Users\damia\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\estimator\training.py", line 886, in evaluate_and_export
    hooks=self._eval_spec.hooks)
  File "C:\Users\damia\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\estimator\estimator.py", line 453, in evaluate
    input_fn, hooks, checkpoint_path)
  File "C:\Users\damia\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\estimator\estimator.py", line 1348, in _evaluate_build_graph
    features, labels, model_fn_lib.ModeKeys.EVAL, self.config)
  File "C:\Users\damia\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\estimator\estimator.py", line 1107, in _call_model_fn
    model_fn_results = self._model_fn(features=features, **kwargs)
  File "C:\Users\damia\Desktop\Magisterka\SSD\models\research\object_detection\model_lib.py", line 383, in model_fn
    include_metrics_per_category=eval_config.include_metrics_per_category)
  File "C:\Users\damia\Desktop\Magisterka\SSD\models\research\object_detection\eval_util.py", line 629, in get_eval_metric_ops_for_evaluators
    input_data_fields.groundtruth_is_crowd)))
  File "C:\Users\damia\Desktop\Magisterka\SSD\models\research\object_detection\metrics\coco_evaluation.py", line 349, in get_estimator_eval_metric_ops
    first_value_op = tf.py_func(first_value_func, [], tf.float32)
  File "C:\Users\damia\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\ops\script_ops.py", line 384, in py_func
    func=func, inp=inp, Tout=Tout, stateful=stateful, eager=False, name=name)
  File "C:\Users\damia\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\ops\script_ops.py", line 227, in _internal_py_func
    input=inp, token=token, Tout=Tout, name=name)
  File "C:\Users\damia\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\ops\gen_script_ops.py", line 130, in py_func
    "PyFunc", input=input, token=token, Tout=Tout, name=name)
  File "C:\Users\damia\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\framework\op_def_library.py", line 787, in _apply_op_helper
    op_def=op_def)
  File "C:\Users\damia\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\framework\ops.py", line 3414, in create_op
    op_def=op_def)
  File "C:\Users\damia\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\framework\ops.py", line 1740, in __init__
    self._traceback = self._graph._extract_stack()  # pylint: disable=protected-access

InvalidArgumentError (see above for traceback): TypeError: can't pickle dict_values objects
Traceback (most recent call last):

  File "C:\Users\damia\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\ops\script_ops.py", line 158, in __call__
    ret = func(*args)

  File "C:\Users\damia\Desktop\Magisterka\SSD\models\research\object_detection\metrics\coco_evaluation.py", line 339, in first_value_func
    self._metrics = self.evaluate()

  File "C:\Users\damia\Desktop\Magisterka\SSD\models\research\object_detection\metrics\coco_evaluation.py", line 193, in evaluate
    self._detection_boxes_list)

  File "C:\Users\damia\Desktop\Magisterka\SSD\models\research\object_detection\metrics\coco_tools.py", line 118, in LoadAnnotations
    results.dataset['categories'] = copy.deepcopy(self.dataset['categories'])

  File "C:\Users\damia\Anaconda3\envs\tensorflow\lib\copy.py", line 174, in deepcopy
    rv = reductor(4)

TypeError: can't pickle dict_values objects


         [[Node: PyFunc_1 = PyFunc[Tin=[], Tout=[DT_FLOAT], token="pyfunc_3", _device="/job:localhost/replica:0/task:0/device:CPU:0"]()]]
         [[Node: SecondStagePostprocessor/BatchMultiClassNonMaxSuppression/MultiClassNonMaxSuppression/non_max_suppression_4/NonMaxSuppressionV3/_3221 = _Recv[client_terminated=false, recv_device="/job:localhost/replica:0/task:0/device:GPU:0", send_device="/job:localhost/replica:0/task:0/device:CPU:0", send_device_incarnation=1, tensor_name="edge_4536_...pressionV3", tensor_type=DT_INT32, _device="/job:localhost/replica:0/task:0/device:GPU:0"]()]]

[manipopopo]

You could change the call to eval_util.get_eval_metric_ops_for_evaluators in model_lib.py
from

eval_metric_ops = eval_util.get_eval_metric_ops_for_evaluators(
          eval_metrics,
          category_index.values(),
          eval_dict,
          include_metrics_per_category=eval_config.include_metrics_per_category)

to

eval_metric_ops = eval_util.get_eval_metric_ops_for_evaluators(
          eval_metrics,
          list(category_index.values()),
          eval_dict,
          include_metrics_per_category=eval_config.include_metrics_per_category)

duplicate: #4780

[programowalny]

Thank you, it work.

[swg209]

@programowalny
Hi, I'd like to know whether the warning is solved using the way above? I take the way but still have this warning ,without TypeError: can't pickle dict_values objects .
WARNING:tensorflow:Ignoring ground truth with image id 65457264 since it was previously added
WARNING:tensorflow:Ignoring detection with image id 65457264 since it was previously added

[programowalny]

I still had the this warning, but i read, You can ignore them

[swg209]

@programowalny But this warning make my model can't get evaluate result? Do you have any idea to sovle?

[programowalny]

What do you mean by model can't get evaluate result? After learning proces you get some errors or your loss function is too high. I still have problem with loss function, after first 5 hundrets step is decreases, but after it grows. And my final model can't recognize any object. Do think this warnings can be a cause of it? Now i changing my enviroment, including OS. Tomorrow i let's you known abaut results

[swg209]

I use model_main.py to train model fit my dataset, and I set every 5min save a ckpt and trigger a evaluation on the model, then I can choose a best model. But the screen keep printing this warning ,
and doesn't show evaluation result.

[programowalny]

You mean that kind of evaluation result? :

[swg209]

Do you use legacy >train.py to train model. The evalution result like this:
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /home/swg/tfmodels/research/object_detection/demo/trains/model.ckpt-9168
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Evaluation [1/5]
INFO:tensorflow:Evaluation [2/5]
INFO:tensorflow:Evaluation [3/5]
INFO:tensorflow:Evaluation [4/5]
INFO:tensorflow:Evaluation [5/5]
creating index...
index created!
INFO:tensorflow:Loading and preparing annotation results...
INFO:tensorflow:DONE (t=0.00s)
creating index...
index created!
Running per image evaluation...
Evaluate annotation type bbox
DONE (t=0.01s).
Accumulating evaluation results...
DONE (t=0.01s).
Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.452
Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 1.000
Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=100 ] = 0.203
Average Precision (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = -1.000
Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.452
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 1 ] = 0.500
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 10 ] = 0.500
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.500
Average Recall (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = -1.000
Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.500

[programowalny]

Ok i tried model_main and train. And i must say i always get the output . When i use model_main i get average precision, when a train.py a loss function. If you did't get a average precision it is not cause of warnings like this:
WARNING:tensorflow:Ignoring ground truth with image id 65457264 since it was previously added

[Victorsoukhov]

I have the same problem. And it was solved by setting
num_examples: 1
in eval_config section of the pipeline config file. As was pointed in the documentation -
[https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/configuring_jobs.md] and as you can see in proto files
[https://github.com/tensorflow/models/blob/master/research/object_detection/protos/eval.proto]
num_examples - is the batch size of the evaluation.

[swg209]

@Victorsoukhov
Thanks a lot. Your way does solve my issue.And I find that the num_example can be set bigger, eg: 10 also works.

[wt-huang]

Closing as this is resolved